coast information team - british columbia · coast information team ... rick tingey, ken...

Coast Information Team

c/o Cortex Consultants Inc., 3A–1218 Langley St. Victoria, BC, V8W 1W2 Tel: 250-360-1492 / Fax: 250-360-1493 / Email: [email protected]

April 30, 2004

The Coast Information Team is pleased to deliver the final version of the CIT Ecosystem Spatial Analysis (April 2004).

The Coast Information Team (CIT) was established to provide independent information for the central and north coasts of British Columbia and Haida Gwaii/Queen Charlotte Islands using the best available scientific, technical, traditional and local knowledge. The CIT was established by the Provincial Government of British Columbia, First Nations, environmental groups, the forest industry, and communities. It is led by a management committee consisting of representatives of these bodies; and is funded by the Provincial Government, the environmental groups and forest products companies, and the Federal Government of Canada. The technical team comprises nine project teams consisting of scientists, practitioners, and traditional and local experts. CIT information and analyses, which include this CIT Ecosystem Spatial Analysis, are intended to assist First Nations and the three ongoing sub-regional planning processes to make decisions that will achieve ecosystem-based management (as per the April 4th 2001 Coastal First Nations—Government Protocol and the CCLRMP Interim Agreement).

In keeping with the CIT’s commitment to transparency and highly credible independent analysis, the CIT Ecosystem Spatial Analysis underwent an internal peer review and the CIT’s independent peer review process chaired by University of Victoria Professor Rod Dobell. Peer reviews of the draft document and the authors’ response are found at http://www.citbc.org/abostru-comm.html. The final document reflects changes made by the authors to address peer review comments.

We encourage all stakeholders involved in land and resource management decision-making in the CIT area to use the information and recommendations/conclusions of the CIT Ecosystem Spatial Analysis in conjunction with other CIT products as they seek to implement EBM and develop EBM Land Use Plans. We are confident that the suite of CIT products provides valuable information and guidance on the key tenets of EBM: maintaining ecosystem integrity and improving human wellbeing.

Sincerely,

Robert Prescott-Allen, Executive Director on behalf of the CIT Management Committee: Ken Baker, Art Sterritt, Dallas Smith, Jody Holmes, Corby Lamb Graem Wells, Gary Reay, Hans Granander, Tom Green, Bill Beldessi


ESA Ltr of Transmittal_mar04


AN ECOSYSTEM SPATIAL ANALYSIS FOR HAIDA GWAII, CENTRAL COAST, AND NORTH COAST BRITISH COLUMBIA

Chuck Rumsey, Jeff Ardron, Kristine Ciruna, Tim Curtis, Zach Ferdaña, Tony Hamilton, Kim Heinemeyer, Pierre Iachetti, Richard Jeo, Gary Kaiser, Debbie Narver, Reed Noss, Dennis Sizemore, Art

Tautz, Rick Tingey, Ken Vance-Borland

April 2004


An Ecosystem Spatial Analysis for Haida Gwaii, Central Coast, and North Coast BC April 2004

Page ii

ACKNOWLEDGMENTS

CIT ESA Planning Team

Dr. Reed Noss (Conservation Science Inc.) mentored the planning team, guided project design and focal species modelling, and contributed greatly to the final written report. Charles Rumsey (Nature Conservancy of Canada) managed the planning team and coordinated analyses, interpreted results, and designed mapping products. Rick Tingey (Round River Conservation Studies) and Ken Vance-Borland (Conservation Science Inc.) were the primary GIS analysts responsible for compiling data sets from all of the technical teams, and inputting and running the SITES and MARXAN models. Dr. Richard Jeo (Round River Conservation Studies), working with Rick, led the creation of the terrestrial coarse filter classification system, the impacts analysis, and the creation of focal species models and interpretation of results. Dr. Kristine Ciruna (Nature Conservancy of Canada) acted as the lead for the freshwater technical team and was responsible for creating the freshwater coarse filter classification system and contributed greatly to editing the final report. Kristine also worked with Dr. Art Tautz (BC Ministry of Water, Land and Air Protection [BC MWLAP], Biodiversity Branch) and Dr. Blair Holtby (Department of Fisheries and Oceans Canada) to create the salmon status and trends models and analyses. Kristine also worked with Tim Curtis (Nature Conservancy of Canada), another GIS analyst on the project, to create the tailed frog focal species model and major components of the salmon analyses. Tim also assisted with many of the final SITES analyses.

Dr. Kim Heinemeyer (Round River Conservation Studies) acted as the focal species technical team lead and was also involved in creating the mountain goat winter range and black-tailed deer winter range modelling with Rick Tingey. Kim and Pierre Iachetti (Nature Conservancy of Canada) coordinated the peer review and revisions to the focal species models. Pierre was the ESA data manager in conjunction with Debbie Narver (BC Ministry of Sustainable Resource Management [BC MSRM], Resource Planning), coordinated the fine filter analysis, acted as liaison between the ESA and many of the agencies and individuals on the project, and contributed to writing and editing of the final report. Debbie also coordinated the habitat modelling for some of the focal species and the creation of some of the data sets used in the ESA such as the amalgamated forest cover. Tony Hamilton (BC MWLAP, Biodiversity Branch) was responsible for much of the coordination of the focal species modelling and acted as a liaison between the ESA and the LRMP technical teams. Tony was also responsible for creation of the grizzly and black bear focal species models. Zach Ferdaña (The Nature Conservancy of Washington) led the nearshore analysis. Zach was responsible for the creation of the shorezone classification system (based on the BC Shorezone classification system) applied to the CIT study area. Jeff Ardron (Living Oceans Society) led the offshore analysis, was responsible for running MARXAN, providing the results of the analysis, and contributed to the writing of this report. Gary Kaiser (Nature Conservancy of Canada) worked on the fine filter analysis and the marbled murrelet focal species model. Gary also provided valuable advice on selecting bird targets for fine and coarse filter analyses.



Page iii

Advisors

Andy MacKinnon (BC MSRM) acted as a liaison between the ESA and LRMP technical teams and provided valuable advice on project design, and the fine and coarse filter analyses. Marvin Eng, Del Meidinger, Allen Banner, and Jim Pojar (all with BC Ministry of Forests) provided advice on the BC biogeoclimatic classification system, small scale Predictive Ecosystem Mapping (SSPEM), and fine filter plant and rare plant communities data sets. Dr. Lance Craighead (Craighead Environmental Research Institute) provided advice and direction on focal species modelling. Dr. Blair Holtby (Department of Fisheries and Oceans Canada) worked with the salmon technical team and provided direction, advice, and data for the analysis. Frank Doyle (Wildlife Dynamics) and Erica McClaren (BC MWLAP) provided direction, advice, and data for the northern goshawk focal species modelling. Carol Ogborne (BC MSRM, Decision Support Services) provided a number of key coastal data sets as well as guidance on the shorezone classification. Alan Burger (University of Victoria and Marbled Murrelet Recovery Team) and Louise Blight (BC MWLAP and Marbled Murrelet Recovery Team) provided advice on building the marbled murrelet focal species model. Pierre Friele and Linda Dupuis (Ascaphus Consulting Inc.) were advisors in the development of the tailed frog focal species model. Steven Northway (Weyerhaeuser and Haida Gwaii/QCI LRMP) provided advice and guidance on the terrestrial coarse filter classification and was able to create a decision matrix for missing Canopy Closure classes in the forest cover data set that allowed us to run the Marbled Murrelet and Northern Goshawk models. Thank you to the LRMP technical teams who provided valuable input to the ESA process and made their staff and data available to the planning team.

Reviewers

Many thanks go to the reviewers of the ESA analyses and products. Sharon Hartwell, Dr. George Douglas, Carmen Cadrin, Andy Stewart, Marta Donovan, and Leah Ramsay (all with BC MSRM, Conservation Data Centre) all provided advice, expertise, and/or data for the fine filter and coarse filter analyses. Dave Nagorsen, Dr. Richard Hebda and Dr. Chris Brayshaw (all with Royal BC Museum) provided advice and review of the plants and animals fine filter targets lists. Todd Golumbia, Pat Bartier, and Dr. Norm Sloan (all with Parks Canada) provided advice as well as made available Gwaii Haanas National Park Reserve data sets to the nearshore and offshore analysis teams.

Thank you to the focal species models reviewers who provided valuable review of the focal species models within tight deadlines. Matt Kirchhoff (Alaska Department of Fish and Game), Dr. Christopher Farmer (consultant), and Kim Brunt (BC MWLAP) reviewed the black-tailed deer winter range model. Grant MacHutchon (consultant), Wayne McCrory (Valhalla Wilderness Society), Dr. John Schoen (Audubon Alaska), Dr. Lance Craighead (Craighead Environmental Research Institute), and Baden Cross (Society for Applied Conservation GIS) reviewed the grizzly and black bears models. Todd Golumbia (Parks Canada) reviewed the black bear model. Monica Mather reviewed the marbled murrelet model. Ken Dunsworth (BC MWLAP), Pierre Friele, and Linda DuPuis (Ascaphus Consulting, Inc.) reviewed the tailed frog model. Erica McClaren (BC MWLAP) reviewed the northern goshawk model. Steve Gordon, Kim Brunt (both with BC MWLAP), Laurence Turney (Ardea Biological Consulting), and Steven Wilson (EcoLogic Research) reviewed the mountain goat winter range modelling.



Page iv

For the offshore marine analysis, Dr. Satie Airame, Dr. Barbara Dugelby, Dr. Reed Noss, Dr. Ian Perry, Dr. Callum Roberts, and Dr. John Roff all provided valuable peer review of Living Oceans Society’s Central Coast analysis, which directly informed the present CIT work. For taking the time to answer data questions and review preliminary results, thanks to Krista Amey, Jacqueline Booth, Kim Conway, Gary Kaiser, Kathleen Moore, and Carol Ogborne.

For the nearshore analysis, thank you to the following people who attended expert workshops and provided valuable advice and guidance: Brenda Bauer (DFO), Carol Ogborne (MSRM), Dora Repard (Canadian Parks & Wilderness Society), Dr. John Harper (Coastal and Ocean Resources Inc.), Dr. Mark Zacharias (California State University Channel Islands), Mary Morris (Archipelago Marine Research, Ltd.), Melody Farrell (Department of Fisheries & Oceans Canada), Natalie Ban (Canadian Parks & Wilderness Society), Pete Wills (Canadian Hydrographic Service/ Department of Fisheries & Oceans Canada), Rob Paynter (BC MSRM), Dr. Rosaline Canessa (AXYS Environmental Consulting/BC MSRM), and Dr. Tom Tomascik (Parks Canada).

Finally, for allowing us to use MARXAN, and for their ongoing technical support: Dr. Ian Ball and Dr. Hugh Possingham.

CIT Management Team and Secretariat

Thank you to the CIT Management Team for their ongoing support of the ESA. The Management Team is co-chaired by Ken Baker (ADM BC MSRM) representing the provincial government and Art Sterritt (Gitga’at First Nation) representing First Nations. Also on the management team are Graem Wells representing local communities and governments, Dr. Jody Holmes (Rainforest Solutions Project) representing environmental non-government organizations (ENGOs), Rick Jeffrey (Truck Loggers Association) representing the forest industry, Dallas Smith (Central Coast LRMP and Tlowitsis First Nation) representing First Nations, and Gary Reay (BC MSRM) representing the provincial government.

Thank you as well to the CIT Secretariat who provided day-to-day support and guidance with the ESA process. Executive Director Robert Prescott-Allen (PDATA) leads the Secretariat. Melissa Hadley (Cortex Consultants) is the Program Manager; Jordan Tanz (Cortex Consultants) is the alternate Program Manager. Edward Gin (Cortex Consultants provided accounting and logistical support to the Secretariat.



Page v

Table of Contents

ACKNOWLEDGMENTS..................................................................................................... ii

EXECUTIVE SUMMARY....................................................................................................1 Background.................................................................................................................................. 1 Regional Classification................................................................................................................... 2 Terrestrial Targets ........................................................................................................................ 3

Terrestrial Special Elements ................................................................................................ 3 Terrestrial Ecosystem Representation .................................................................................. 3 Terrestrial Focal Species ..................................................................................................... 4

Freshwater Targets....................................................................................................................... 8 Freshwater Special Elements ............................................................................................... 8 Freshwater Representation.................................................................................................. 8 Freshwater Focal Species .................................................................................................... 9

Nearshore Marine Targets ........................................................................................................... 11 Introduction ..................................................................................................................... 11 Nearshore Marine Special Elements ................................................................................... 11 Nearshore Marine Ecosystem Representation ..................................................................... 12

Offshore Marine Targets.............................................................................................................. 12 Special Elements and Focal Species ................................................................................... 12 Offshore Marine Ecosystem Representation........................................................................ 13

Setting Goals.............................................................................................................................. 14 Threats (Human Impacts) Analysis .............................................................................................. 14 Spatial Analysis........................................................................................................................... 15

Terrestrial and Freshwater Analysis ................................................................................... 15 Nearshore Marine Analysis ................................................................................................ 16 Offshore Marine Analysis................................................................................................... 17 Options and Scenarios ...................................................................................................... 17 Alternative Options and Scenarios for Conservation ............................................................ 18 Relative Biodiversity Index ................................................................................................ 19

Spatial Analysis Results ............................................................................................................... 19 Terrestrial/Freshwater Spatial Analysis Results ................................................................... 19 CC LRMP Representation Analysis ...................................................................................... 20 Nearshore Marine Spatial Analysis Results.......................................................................... 20 Offshore Marine Spatial Analysis Results ............................................................................ 21

Conclusions ................................................................................................................................ 21



Page vi

1.0 INTRODUCTION...................................................................................................23 1.1 The Planning Process ........................................................................................................ 25 1.2 Study Area and Objectives ................................................................................................ 25

1.2.1 Study Area ............................................................................................................. 25 1.2.2 Haida Gwaii, Central Coast, and North Coast British Columbia ................................... 26 1.2.3 Biological Stratification............................................................................................ 27 1.2.4 The ESA Study Area Boundaries .............................................................................. 28 1.2.5 Study Area Ecoregions and Ecosections ................................................................... 29 1.2.6 ESA Ecological Drainage Units ................................................................................. 30 1.2.7 Study Area Ecological Description ............................................................................ 31 1.2.8 The Coast Information Team................................................................................... 32 1.2.9 The CIT Ecosystem Spatial Analyses ........................................................................ 34

2.0 CONSERVATION TARGETS ...................................................................................36 2.1 The Three Track Approach to Setting Conservation Targets and Goals ................................. 36

2.1.1 Special Elements .................................................................................................... 36 2.1.2 Representation ....................................................................................................... 37 2.1.3 Focal Species ......................................................................................................... 38

2.2 Terrestrial Targets ............................................................................................................ 38 2.2.1 Special Elements .................................................................................................... 38 2.2.2 Representation ....................................................................................................... 42 2.2.3 Focal Species: Marbled Murrelet Nesting Habitat Suitability ....................................... 48 2.2.4 Focal Species: Northern Goshawk Nesting Area and Foraging Area Habitat Suitability

Models................................................................................................................... 52 2.2.5 Focal Species: Black-Tailed Deer Winter Range Model............................................... 56 2.2.6 Focal Species: Mountain Goat (Oreamnos americanus) Winter Habitat Model ............. 58 2.2.7 Focal Species: Grizzly Bear (Ursus arctos) Habitat Suitability and Capability Models .... 62 2.2.8 Focal Species: Black Bear (Ursus americanus) Habitat Suitability and Capability Model 67

2.3 Freshwater Targets........................................................................................................... 68 2.3.1 Special Elements .................................................................................................... 68 2.3.2 Representation Analysis .......................................................................................... 68 2.3.3 Focal Species: Tailed Frog (Ascaphus truei) Habitat Suitability Model ......................... 73 2.3.4 Focal Species: Pacific Salmon and Steelhead ............................................................ 76

2.4 Marine Targets ................................................................................................................. 93 2.4.1 Nearshore Marine ................................................................................................... 93 2.4.2 Offshore Marine Analysis....................................................................................... 101

3.0 SETTING GOALS.................................................................................................108 3.1 Background.................................................................................................................... 108

3.1.1 How Much Is Enough: Level of Landscape Protection (Representation Analysis) ....... 108 3.1.2 Goals for Focal Species ......................................................................................... 109 3.1.3 Conclusions.......................................................................................................... 110

3.2 CIT ESA Terrestrial/Freshwater Goal Setting..................................................................... 112



Page vii

3.3 Goals for the Marine Nearshore Environment.................................................................... 112 3.3.1 Representation (Coarse Filter) Goals ...................................................................... 112

3.4 Goals for Offshore Marine ............................................................................................... 117

4.0 HUMAN IMPACTS...............................................................................................118 4.1 Background and Rationale............................................................................................... 118 4.2 Methods......................................................................................................................... 119 4.3 Results........................................................................................................................... 119

5.0 SPATIAL ANALYSIS: METHODS..........................................................................122 5.1 Background.................................................................................................................... 122 5.2 Spatial Analysis: Design Process and Tools....................................................................... 122

5.2.1 Steps of Spatial Analysis ....................................................................................... 122 5.2.2 Automated Site Selection Algorithms...................................................................... 123 5.2.3 Terrestrial and Freshwater Spatial Analysis............................................................. 124 5.2.4 Nearshore Marine Spatial Analysis ......................................................................... 126 5.2.5 Marine Deep-water Spatial Analysis ....................................................................... 132

5.4 Options and Scenarios .................................................................................................... 136 5.4.1 Background.......................................................................................................... 136 5.4.2 Conservation Utility............................................................................................... 136 5.4.3 Condition ............................................................................................................. 137 5.4.4 Alternative Options for Conservation...................................................................... 137 5.4.5 Alternative Scenarios Analysis................................................................................ 138 5.4.6 Comparing Options and Scenarios ......................................................................... 138 5.4.7 Relative Biodiversity Index (RBI) ........................................................................... 138

6.0 SPATIAL ANALYSIS: RESULTS AND DISCUSSION .............................................139 6.1 Terrestrial/Freshwater Spatial Analysis Results ................................................................. 139

6.1.1 Options and Scenarios measured by Analysis Unit................................................... 139 6.1.2 Protected Areas: Existing and Potential .................................................................. 141 6.1.3 Relative Biodiversity Index .................................................................................... 142 6.1.4 Representation Analysis of CC LRMP Negotiated Agreement.................................... 143

6.2 Marine Nearshore Spatial Analysis Results ........................................................................ 145 6.3 Offshore Marine Spatial Analysis Results .......................................................................... 152

6.3.1 24 Scenarios; 2,400 Solutions................................................................................ 152 6.3.2 Utility................................................................................................................... 152 6.3.3 Flexible Solutions.................................................................................................. 154

7.0 NEXT STEPS AND FUTURE WORK ......................................................................155 7.1 Conservation Assessment of LRMP and LUP and Government-to-Government Decisions...... 155 7.2 Special Elements Refinement........................................................................................... 155 7.3 Integrated Marine, Terrestrial, Freshwater ....................................................................... 155 7.4 Alignment of Terrestrial Classification Systems ................................................................. 155 7.5 Enhancements to Salmon Modelling................................................................................. 156



Page viii

7.6 Species Penalties/Variable Goals ...................................................................................... 156 7.7 Temporal Modelling ........................................................................................................ 156 7.8 Cross-Scale Modelling ..................................................................................................... 156

8.0 CONCLUSIONS ...................................................................................................157

9.0 LITERATURE CITED............................................................................................159

GLOSSARY ..................................................................................................................185

MAPS (attached separately at http://citbc.org/anaecos.html).................................187 Map 1. Central Coast, North Coast, and Haida Gwaii LRMP Boundaries......................................... 187

Map 2. Ecosections and Terrestrial Stratification Units................................................................. 187 Map 3. Ecological Drainage Units ............................................................................................... 187 Map 4. Conservation Data Centre Element Occurrence Records ................................................... 187 Map 5. Forested and Developed Landscape ................................................................................ 187 Map 6a. Biogeoclimatic Ecosystem Classification ......................................................................... 187 Map 6b. Seral Stage and Site Index ........................................................................................... 187 Map 7. Focal Ecosystems........................................................................................................... 187 Map 8. Habitat Suitability Model for Marbled Murrelet ................................................................. 187 Map 9. Nesting and Foraging Habitat Suitability Models for Northern Goshawk ............................. 187 Map 10. Mountain Goat Winter Range Habitat Model .................................................................. 187 Map 11. Grizzly Bear and Black Bear Habitat Capability ............................................................... 187 Map 12. Freshwater Ecosystems................................................................................................ 187 Map 13. Habitat Suitability Model for Tailed Frog ........................................................................ 187 Map 14. Steelhead Population Status: Winter Run ...................................................................... 187 Map 15. Steelhead Population Status: Summer Run.................................................................... 187 Map 16. Salmon Population Status: Coho ................................................................................... 187 Map 17. Salmon Population Status: Chinook............................................................................... 187 Map 18. Salmon Population Status: Chum .................................................................................. 187 Map 19. Salmon Population Status: Even Year Pink..................................................................... 187 Map 20. Salmon Population Status: Odd Year Pink...................................................................... 187 Map 21. Salmon Population Status: Sockeye............................................................................... 187 Map 22. Nearshore Ecosection Stratifications.............................................................................. 187 Map 23. Nearshore Cost Parameters .......................................................................................... 187 Map 23a. Shoreline Cost Index .................................................................................................. 187 Map 24. Nearshore Data Collection Regions................................................................................ 187 Map 25. Seabird Colonies .......................................................................................................... 187 Map 26. Moulting Seaducks ....................................................................................................... 187 Map 27. Anadromous Streams................................................................................................... 188 Map 28. Bathymetry ................................................................................................................. 188



Page ix

Map 29. Benthic Topographical Complexity................................................................................. 188 Map 30. Ecological Regions and Data Regions ............................................................................ 188 Map 31. Human Impacts Assessment......................................................................................... 188 Map 32. Conservation Utility of CIT Analysis Units: Existing protected areas locked in................... 188 Map 33. Conservation Utility of CIT Analysis Units: Existing and

Candidate protected areas locked in. ............................................................................... 188 Map 34. Conservation Utility of CIT Analysis Units: Existing, Candidate, and

Option protected areas locked in. .................................................................................... 188 Map 35. Conservation Utility of CIT Analysis Units: No areas locked in. ........................................ 188 Map 36a. Tier One Analysis Units for CIT ESA: Existing protected areas locked in. ........................ 188 Map 36b. Conservation Priority for CIT ESA Analysis Units: Existing protected areas locked in. ...... 188 Map 37a. Tier One Analysis Units for CIT ESA: Existing and Candidate protected areas locked in... 188 Map 37b. Conservation Priority for CIT ESA Analysis Units: Existing and

Candidate protected areas locked in ................................................................................ 188 Map 38a. Tier One Analysis Units for CIT ESA: Existing, Candidate and

Option protected areas locked in. .................................................................................... 188 Map 38b. Conservation Priority for CIT ESA Analysis Units: Existing, Candidate and

Option protected areas locked in. .................................................................................... 188 Map 39a. Tier One Analysis Units for CIT ESA: No areas locked in. .............................................. 188 Map 39b. Conservation Priority for CIT ESA Analysis Units: No areas locked in. ............................ 188 Map 40. Relative Biodiversity Index for Coastal LRMP’s. .............................................................. 188 Map 41. Neashore Summed Solution Gradient ............................................................................ 188 Map 42. Nearshore Option A ..................................................................................................... 188 Map 43. Nearshore Option B ..................................................................................................... 188 Map 44. Nearshore Option C ..................................................................................................... 188 Map 45. Deepwater Marine Conservation Utility .......................................................................... 188

Appendices (attached separately at http://citbc.org/anaecos.html) .......................189



Page x

List of Figures Figure 2.1 Summary of salmon species condition within CIT study area............................................ 84

Figure 2.2 Summary of coho condition by EDU................................................................................ 85

Figure 2.3 Summary of chum condition by EDU............................................................................... 86

Figure 2.4 Summary of sockeye condition by EDU. .......................................................................... 87

Figure 2.5 Summary of odd year pink condition by EDU................................................................... 88

Figure 2.6 Summary of even year pink condition by EDU. ................................................................ 89

Figure 2.7 Summary of chinook condition by EDU. .......................................................................... 90

Figure 2.8 Summary of summer steelhead condition by EDU............................................................ 91

Figure 2.9 Summary of winter steelhead condition by EDU. ............................................................. 92

Figure 6.1 Conservation tiers for CIT ESA analysis units, comparing protected area scenarios. ......... 140

Figure 6.2 Progress toward representation goals for ESA SITES summed solutions: Existing protected areas locked in. ............................................................................... 141

Figure 6.3 Proportion of CIT ESA Conservation targets represented in existing and proposed protected area scenarios based on 5 example representation thresholds......................... 142

Figure 6.4 Focal Species habitat representation based on CC LRMP agreement. .............................. 144

List of Tables Table 2.1 Special elements target selection criteria ........................................................................ 40

Table 2.2 Summary of terrestrial fine filter targets ......................................................................... 41

Table 2.3 Biogeoclimatic zones and logging impacts by site productivity .......................................... 45

Table 2.4 Ecological system and alliance species groupings based on ITG ....................................... 47

Table 2.5 Focal ecological system structure classes ........................................................................ 48

Table 2.6 Model structure for Marbled Murrelet habitats in the CIT study area................................. 51

Table 2.7 Key structural attributes and forest species composition of goshawk nesting habitat ......... 53

Table 2.8 Indicators of nest area suitability.................................................................................... 53

Table 2.9 Indicators of foraging area suitability.............................................................................. 54

Table 2.10 Black-tailed deer winter range model parameters and values ........................................... 57

Table 2.11 Landscape features and assumptions used in mapping for mountain goat winter range..... 60

Table 2.12 CIT area-wide mountain goat potential winter model definitions and assumptions............. 60

Table 2.13 Summary of data used in coarse-scale freshwater ecosystem classification ....................... 71

Table 2.14. Summary of coarse-scale freshwater system types by EDU. ............................................. 72

Table 2.15 Summary of weightings used to set conservation goals for all salmon SPUs ...................... 79

Table 2.16 CIT regional summary of salmon status .......................................................................... 81

Table 2.17 Summary of salmon status for Nass River EDU................................................................ 81

Table 2.18 Summary of salmon status for Skeena River EU .............................................................. 82

Table 2.19 Summary of salmon status for North Coast EDU.............................................................. 82



Page xi

Table 2.20 Summary of salmon status for Central Coast EDU ........................................................... 82

Table 2.21 Summary of salmon status for Haida Gwaii EDU.............................................................. 83

Table 2.22 BC Coastal classes and representative types ................................................................... 96

Table 2.23 Shoreline ecosystem targets .......................................................................................... 97

Table 2.24 Summary of data layers by type................................................................................... 102

Table 3.1 Percentage of land recommended for protection in a number of regions......................... 110

Table 3.2 Shoreline ecosystem goals ........................................................................................... 113

Table 3.3 Vegetation and fine filter goals..................................................................................... 116

Table 3.4 Actual percentages attached to each qualitative ranking ................................................ 117

Table 4.1 Reported thresholds for human impacts on biodiversity ................................................. 120

Table 4.2 ESA human impact classes........................................................................................... 120

Table 4.3 Watersheds and river systems...................................................................................... 121

Table 5.1 CIT ESA conservation utility ratings based on SITES summed solution scores ................. 137

Table 5.2 Conservation tiers for the CIT ESA analysis units........................................................... 137

Table 6.1 Comparison among option and scenario combinations for CIT ESA analysis units ............ 139

Table 6.2 Protected area status of analysis units from the “No areas locked in” SITES solution. ...... 142

Table 6.3 Coarse filter targets captured by nearshore Options A, B, and C..................................... 147

Table 6.4 Fine Filter targets captured by nearshore Options A, B, and C........................................ 151



Page xii



Page 1

EXECUTIVE SUMMARY

Background

This report presents a comprehensive ecosystem spatial analysis for the Haida Gwaii, Central Coast, and North Coast of British Columbia. This region has a land area of 11 million hectares (ha); its sea area is another 11 million ha. Over 95% or the land area is designated Crown land and is managed by the Government of British Columbia. In addition to productive, structurally diverse old-growth ecosystems and unique bog complexes, important ecological elements in the region include unregulated rivers supporting large populations of spawning salmon and grizzly bears, estuaries, kelp beds, seabird colonies, archipelago/fjord terrain, deep fjord and cryptodepression lakes, and intertidal flats with abundant invertebrates and resident and migratory waterbirds. Haida Gwaii is an especially significant part of the region, containing an insular biota with distinctive, disjunct, and endemic taxa. The diversity of species within the CIT region is far greater than previously thought, but still incompletely known.

Two major land use and resource-management planning processes (LRMPs) are underway in the region: the Central Coast LRMP and the North Coast LRMP. The Haida Gwaii/Queen Charlotte Islands Plan is in development (Map 1). Their purpose is to enable all parties to reach agreement on those lands and resources to be protected and those to be developed, where, and how. The Government of British Columbia, First Nations, environmental groups, and forest products companies established the Coast Information Team (CIT) to provide independent information on the region using the best available scientific, technical, traditional, and local knowledge. The CIT’s information and analyses, which include this ecosystem spatial analysis, are intended to assist First Nations and the three ongoing subregional planning processes to make decisions that will achieve ecosystem-based management (EBM).

The purpose of the ecosystem spatial analysis is to identify priority areas for biodiversity conservation and, to provide an information base and decision support for subsequent planning and management efforts designed to address four well-accepted goals of conservation: 1) represent ecosystems across a range of environmental gradients; 2) maintain viable populations of native species; 3) sustain ecological and evolutionary processes within a natural range of variability; and 4) build a conservation network that is resilient to environmental change. In pursuit of these goals, the ESA integrates three basic approaches to conservation planning:

• Representation of a broad spectrum of environmental variation (e.g., vegetation, terrestrial abiotic, and freshwater and marine habitat classes).

• Protection of special elements: concentrations of communities; rare or at-risk communities; rare physical habitats; concentrations of species; locations of at-risk species; locations of highly valued species or their habitats; locations of major genetic variants.

• Conservation of habitats for focal species, whose needs help planners address issues of habitat area, configuration, and quality. These species (a) need large areas or several well connected areas, or (b) are sensitive to human disturbance, and (c) can have habitat suitability models available or constructed for them.



Page 2

This approach does not explicitly examine the trade-offs between protected areas and management regimes applied to the intervening landbase or “matrix.” Ultimately, the information provided by the ESA on priority conservation areas must be integrated into the CIT’s proposed Ecosystem-based Management (EBM) Framework. Through such a framework there is the opportunity to explore how conservation objectives can be met by balancing management activities along a spectrum of ecological risk.

As a coarse-scale, relatively low-resolution assessment, a regional conservation plan such as the ESA is not intended to offer detailed guidance for site-level management of either protected areas or the landscape matrix. Such guidance is better provided by through ecosystem-based management at the site and landscape levels (e.g., the province’s EBM handbook). The ESA, like other regional conservation assessments, takes a macroscopic view of the region, and is useful for 1) highlighting areas of regional biological significance; 2) portraying the spatial pattern of high-value sites on a broad scale; 3) illuminating the landscape context of these sites; 4) assessing the conservation needs of wide-ranging (i.e., “regional-scale” and “coarse-scale”) species; and 5) identifying priorities for further, more detailed and research on finer spatial scales. Again, regional conservation planning is not intended to address site-level issues, which require high resolution of habitat features. Hence, regional-scale planning is not appropriate for developing conservation guidelines for species, such as many small vertebrates, invertebrates, and plants, whose distributions and viability are determined by habitat features and processes operating at fine spatial scales. For a comprehensive assessment of conservation and management needs, regional-scale planning should be followed by progressively more detailed research and planning at landscape, watershed, and local scales. We explicitly identify, in our chapter on Next Steps and Future Work, the need to integrate the results of the ESA with the work of the proposed coastal Ecosystem-Based Management Council that is being negotiated as part of the LRMP and ongoing government-to-government process. We strongly endorse future analyses at sub-regional, landscape, watershed, and stand scales, which will complement the information developed through the ESA study.

Finally, the ESA team used the best available information for this assessment but recognizes that new and more comprehensive data will continually become available. Therefore, the ESA should be regarded as an initial step in an iterative assessment process.

Regional Classification

One advantage of a regional approach to planning is that it can place any landscape feature in a local, regional, or global context. A second important advantage is that species, plant communities, and other conservation targets can be considered together within an environmental framework that shaped their evolution and continues to shape their interactions. The Ecoregion Classification system in common use in Canada stratifies British Columbia’s terrestrial and marine ecosystem complexity into discrete geographical units at five levels: ecodomains, ecodivisions, ecoprovinces, ecoregions, and ecosections. The lower levels in this classification describe areas of similar climate, physiography, oceanography, hydrology, vegetation, and wildlife potential.

The CIT study area falls within the Coast and Mountains Ecoprovince of BC. This Ecoprovince extends from coastal Alaska to coastal Oregon. In British Columbia it includes the windward side



Page 3

of the Coast Mountains and Vancouver Island, all of the Queen Charlotte Islands, and the Continental Shelf including Dixon Entrance, Hecate Strait, Queen Charlotte Strait, and the Vancouver Island Shelf. The Coast and Mountains Ecoprovince consists of the large coastal mountains, a broad coastal trough and the associated lowlands, islands and continental shelf, as well as the insular mountains on Vancouver Island and the Queen Charlotte Islands archipelago. The study area overlaps with a total of five terrestrial-based ecoregions encompassing 10 ecosections (Map 2). A sixth ecoregion is dominated by the open water marine environment and can be further divided into four marine ecosections.

In addition to using terrestrial-based classifications for stratifying the study area, we delineated broad ecological drainage units (EDUs) as part of a freshwater ecosystem classification framework (Map 3). EDUs are groups of watersheds that share a common zoogeographic history and physiographic and climatic characteristics. EDUs contain sets of freshwater ecosystem types with similar patterns of drainage density, gradient, hydrologic characteristics, and connectivity. Identifying and describing EDUs allows us to stratify the study area into smaller units so we can better evaluate patterns of freshwater community diversity. Additionally, EDUs provide a means to stratify the study area to set conservation goals.

Terrestrial Targets

Terrestrial Special Elements

Special element targets were initially selected based on conservation status at various scales. The initial list of targets was then reviewed by experts and pared down on the basis of this review. Element occurrence data were collected from all known sources, with most from the BC Conservation Data Centre, which collects and maintains data on rare and endangered species and other features in the province. Data were screened based on their quality and reliability. Extinct, extirpated, and historic species occurrences were removed, as were animal occurrences older than 25 years and plant occurrences older than 40 years. The special elements database consists of 110 targets (75 vascular plants, 9 birds, 5 mammals, and 21 rare plant communities). Spatial data was obtained for 72 targets (Map 4).

The special element data set is small, with relatively few actual point and polygon locations. The majority of element occurrence data is on Haida Gwaii/QCI, reflecting an uneven distribution of surveying effort. Although there are concerns regarding biased survey data—e.g., many surveys are close to roads—special elements need to be protected where they are known to occur. To that end, instead of using these data as an input to the site selection algorithm, the ESA has simply packaged occurrence data as part of its final database so as to allow decision-makers and stakeholders at LRMP tables to assess protected areas and management regimes for how well these elements are being represented.

Terrestrial Ecosystem Representation

Two complementary regional ecosystem classification systems exist in BC: 1) the ecoregion/ecosection classification, which describes broad regional ecosystems based on the interaction of climate and physiography; 2) the biogeoclimatic ecosystem classification (BEC), which delineates ecological zones (biogeoclimatic units) by vegetation, soils, and climate. We used both systems and applied two separate classification methods to assess the representation of ecosystems in the study region.



Page 4

The BEC delineates ecological zones (biogeoclimatic units) by vegetation, soils, and climate. We combined BEC information with site productivity (taken from site index in the BC forest cover data), to delineate terrestrial ecosystems (Maps 5, 6a, 6b). We assumed that impacted areas of the same BEC zone and site index as non-impacted areas will have similar old-growth characteristics; we also assumed a relatively low level of natural disturbance. This method also allows us to preferentially identify and represent non-impacted areas, and then represent comparable impacted areas, only if necessary to meet ecosystem representation goals.

Site index was grouped into three classes (Low = 1–14, Medium = 15–21, and High = > 22). BEC zones were taken from the BC biogeoclimatic classification coverage. We treated separate variants and subzones as separate units because in some areas, variants were not defined. Logging data were taken from several sources; when there was a discrepancy, we considered areas logged and set Age Class = 1. Historical abundance of each ecosystem type was calculated as the sum of area for all seral stages.

Goals for the site-selection algorithm were set at 30, 40, 50, 60, and 70% representation of total area for the ecosystem type based on historical abundance. However, for some ecosystem types, the representation goal exceeded the remaining unlogged forest. Representation targets were selected from areas that had both site index and species information (i.e., with known vegetation information), to eliminate bare rock and ice areas. We applied the following rules for goal setting: 1) If a representation goal can be met using non-impacted areas (i.e., target ≤ remaining forest), no goal was set for logged areas; goals for remaining forest were based on historical abundance estimates. 2) If impacts equal or exceed 85%, the remaining old-growth forest and young non-impacted forest goals were set to 100%, and the goal for logged forest was set as the difference between the overall representation target area and the remaining non-impacted forest area using a formula: Logged forest area target (ha) = representation goal/100 * historic distribution area (ha) – (old growth (ha) + non-impacted young forest (ha)). 3) Where impacts were less than 85%, a 95% target was set for intact areas, with remaining representation target set using logged areas as above.

To complement the site productivity and BEC method for identifying ecosystems, we combined floristic and structural data to identify focal ecological systems (Map 7). Several species of conifers occur in the coastal rainforest of BC; we used size class, inventory type group, and age class to define and delineate stands of old growth and woodland areas based on both floristic and structural characteristics. We grouped inventory type groups to identify ecological systems and, because forest ecosystems differ across the region depending on the physical environment and other factors, we stratified goal setting by ecosection. In the absence of specific information on the location of vegetative communities, this method provided us with a coarse filter that allowed for the representation of a full range of ecosystem types. Focal ecosystems were defined as unique combinations of structure, ecological system, and ecosection. Goals for focal ecosystems were varied from 30 to 70% in 10% increments.

Terrestrial Focal Species Marbled Murrelet

The Marbled Murrelet is a small seabird whose biology has given it an important role in resource management in British Columbia. With only a few exceptions, its nests are in old-growth forest.



Page 5

The Marbled Murrelet was chosen as a focal species due to conservation concerns over loss of nesting habitat and increased predation risk from forestry activity. There is also increasing concern for the influences of human activity on its survival at sea.

The habitat suitability model for this species was based on the Conservation Assessment developed for the Marbled Murrelet Recovery Team. Two versions of the model were run. The first version included the habitats “most likely” to be preferred by the murrelets. In the second version, elevation, tree height, and canopy closure classes were broadened to include other habitats “moderately likely” to be preferred by murrelets. Only the first version was used in the site selection process (Map 8).

The model showed a widespread distribution of murrelet habitat without any strong concentrations or significant isolated pockets. Most lowland areas and valley bottoms were included. Sites for known nests were captured in the Haida Gwaii/Queen Charlotte Islands, but many of the known nest sites around Mussel Inlet occur at too high an elevation to be captured. The model indicated the presence of 216,093 ha of the most likely habitat on the Haida Gwaii/Queen Charlotte Islands. No areas were added by relaxing the criteria. There were 359,699 ha on the Central Coast and 154,736 ha on the North Coast that increased by 16,356 ha and 2,858 ha respectively, when the criteria were relaxed. The model provides a general picture of the distribution of murrelet habitat along the coast of BC, but provides no information about variation in quality or usage.

Northern Goshawk

The Northern Goshawk is a broad-winged, long-tailed raptor of medium size adapted to forest habitats. It was chosen as a focal species because of its strong association with mature/old-growth coniferous forests for foraging and nesting, and the possible impacts to these habitats from forestry activities. Within the CIT study area there are two subspecies of Northern Goshawk.

A goshawk breeding territory has three hierarchical components: nest area, post-fledging area, and foraging area. Habitat-suitability models for the nest area and foraging area components of a goshawk territory were developed for the three LRMP areas in the study region (Map 9). The models are based on those developed for the North Coast Forest District and combined expert opinion, observed habitat characteristics at goshawk nest areas in the Coastal Western Hemlock (CWH) biogeoclimatic zone, and relevant literature. Goshawks consistently select key structural attributes and forest species composition for nesting. The foraging area is the entire area used by the adults. Goshawk diets are dominated by prey associated with mature forests, which favours hunting primarily in mature/old-growth forest areas with high canopy closure and a clear understory. The output of the model is a grid with habitat suitability ratings.

Black-Tailed Deer

The black-tailed deer was chosen as a focal species for the Central Coast and North Coast portions of the study area because loss of old-growth forest cover has a high potential to negatively affect deer populations. The complex canopy structure of old-growth forests allows sufficient light to penetrate, promoting the growth of a diverse set of vascular and non-vascular plants for forage, as well as providing for interception of snow. Deer abundances also ultimately affect predator–prey relationships.



Page 6

We targeted black-tailed deer winter range as a key component of their habitat, which acts as a limiting factor in deer abundances. The amount and duration of snowfall an area receives strongly influences its ability to support deer. During periods of deep and prolonged snow, deer look for old, high-volume forests on gentle to moderate slopes at low elevations.

The black-tailed deer winter range model is based on an “old-growth, deer winter-range” model. We applied a GIS to spatially identify deer winter range in the study area. Deer winter range habitat was only modelled for the Central and North Coast portions of the study area.

The model identifies 688,775 ha of black-tailed deer winter range in the study area (517,835 ha in the Central Coast and 170,628 ha in the North Coast). Habitat was distributed evenly throughout the two LRMP areas and as expected, in areas featuring older coastal western hemlock forests including many of the islands. Modifications to the model suggested by reviewers (e.g., the inclusion of aspect, possibly exposure, and the removal of slope steepness) would more accurately reflect black-tailed deer winter range in the study area. However, due to time and capacity constraints, the ESA focal species team was not able to modify and rerun the model based on expert review recommendations. As a result, the model was not used as an input to the site-selection algorithm.

Mountain Goat

The mountain goat occupies steep, rugged terrain in the mountains of northwestern North America. British Columbia is home to up to 60% of the continental mountain goat population. Generally, mountain goats inhabit alpine and subalpine habitats in all of the mountain ranges of the province. The characteristically rugged terrain is comprised of cliffs, ledges, projecting pinnacles, and talus slopes. The availability of winter range may be a limiting factor for goat populations. Winter habitats may be low elevation habitats where snow accumulation is low, or high elevation habitats where wind, sun or precipitous terrain adequately shed snow from foraging habitats. In coastal BC areas, mountain goats generally move to south-facing, forested areas that offer reduced snow loading and increased access to foraging. The mountain goat was selected as a focal species because of its sensitivity to direct impacts of forest cover removal from limited winter ranges, as well as the potential direct and indirect mortality associated with increased access to human activity.

In many regions of the study area, past surveys and fine-scale habitat modelling has identified present populations and associated winter habitats. These data and the modelling results were used to identify the highest priority mountain goat areas in the region. Additionally, a coarse-scale, spatially explicit model was developed to predict potential mountain goat winter habitats in areas not included by these previous efforts, and also to predict potential future habitat potential (i.e., habitat capability). This effort builds on the foundational GIS-based modelling used to identify areas for finer-scale habitat identification.

The combined models identified approximately 300,000 ha of occupied winter goat habitat in the CIT study area. These habitats are not equally distributed across the study area, due to the project-specific nature of the data. The CIT-wide potential winter habitat model overlaps 65% of these occupied winter mountain goat habitats, and predicts additional potential winter habitat across the study area (Map 10).



Page 7

Grizzly Bear

Grizzly bears are found throughout coastal British Columbia, with the exception of the Georgia Depression Ecoprovince, Vancouver Island, Queen Charlotte Islands, and the Coastal Douglas-fir (CDF) biogeoclimatic zone. Coastal grizzly bears are mostly solitary, intra-specifically aggressive omnivores that typically have large seasonal and annual home ranges. They require habitat that provides for their nutritional, security, thermal, reproductive and “space” needs. To meet these varied needs, bears use an array of habitats, ranging from subalpine to valley bottom, old growth to young forest, and wetlands to dry areas. With the exception of denning areas and avalanche chutes, the prime habitat of coastal grizzlies occurs predominantly below treeline and is largely concentrated in valley-bottom ecosystems often associated with important salmon streams. Grizzly bears were chosen as focal species for the CIT ESA because they can be keystone species (transporting salmon away from spawning channels), indicators (because they are susceptible to a wide variety of human influences and have low population densities), and umbrellas (representing a number of species because of their use of such a wide variety of habitats).

The CIT ESA grizzly bear model is a developmental extension of the provincial grizzly bear estimation process commonly referred to as the Fuhr-Demarchi method. The premise behind the approach is that different ecological units vary in their ability to support grizzly bear food resources and that such variations are linked (linearly) to bear density. Instead of attempting to translate habitat effectiveness into bear density, the model used here simply reports grizzly bear habitat effectiveness across the region. The model used the following data as indicators of habitat capability and suitability: 1) broad ecosystem units (BEU); 2) TRIM 1:20,000 Digital Elevation Model (DEM); 3) salmon biomass estimates; and 4) roads/road density.

The primary model outputs are habitat effectiveness ratings for each of the 5,983 modelled watersheds in occupied grizzly bear range within the study area (Map 11). A parallel coverage of target effectiveness classes was also produced. Comparison of the current effectiveness with the proposed target effectiveness can be instructive: watersheds with high correspondence between current and proposed target effectiveness may be high priorities for protection or “light touch” ecosystem-based management. Watersheds with high discrepancy between current estimates and proposed target estimates may be good candidates for modified management, such as through road closures.

Black Bear

Black bears are commonly found in all biogeoclimatic zones in British Columbia. They are very adaptable and inhabit a wide variety of habitats, from coastal estuaries to high elevation alpine meadows. Grasses, sedges, and horsetails form the bulk of their diet, particularly in late spring and early summer. They also feed on insects, fruits, berries, fish, garbage, carrion, small mammals, and occasionally on young deer. In the late summer and fall, salmon spawning rivers and streams represent important feeding areas. Black bear habitat use is strongly influenced by intraspecific social interactions and the presence and activities of people.

The black bear model is also a modification of the Fuhr-Demarchi grizzly bear estimation process. The higher any one land area is ranked for its ability to provide black bear foods, the higher the habitat effectiveness rating attached to it. Black bear habitat was mapped only outside of grizzly bear range, i.e., on Haida Gwaii/QCI and along the mainland coast. Data used in creating the model included 1) BEI – ratings table; 2) TRIM 1:20,000 Digital Elevation Model (DEM); 3) salmon



Page 8

biomass estimates; 4) roads; 5) settlements/towns/recreation user days; and 6) shoreline – rated for seasonal habitat availability for foraging.

One of the objectives of the black bear model is to propose appropriate population targets for conservation of the species under ecosystem-based management. Black bears were also chosen as focal species because they inhabit a wide variety of habitats—from coastal estuaries to high alpine meadows—and in capturing their habitat we assumed that this would also sweep in other species that we might not have data for that use the same habitats. As for grizzly bear, comparison of the current (effectiveness) estimates and densities with the proposed target estimates and densities can be instructive for protected areas design and for management of the landscape matrix. The highest quality habitats were identified on Haida Gwaii/QCI on the west side of Moresby Island and Graham Island (Map 11).

Freshwater Targets

Freshwater Special Elements

Salmonids aside, there was very little spatially explicit special element data available for freshwater species within the study area. Four freshwater fish species were selected for the target list. We were only able to obtain spatial data for the Giant black stickleback, an endemic species that is critically imperiled globally. Three amphibians were selected for the target list but only one had spatial location data (the coastal tailed frog, also a focal species).

Freshwater Representation

Freshwater ecosystems consist of a group of strongly interacting communities held together by shared physical habitat, environmental regimes, energy exchanges, and nutrient dynamics. Freshwater ecosystems are extremely dynamic in that they often change where they exist (e.g., a migrating river channel) and when they exist (e.g., seasonal ponds). Freshwater ecosystems fall into three major groups: standing-water ecosystems (e.g., lakes and ponds); flowing-water ecosystems (e.g., rivers and streams); and freshwater-dependent ecosystems that interface with the terrestrial world (e.g., wetlands and riparian areas).

The classification of freshwater ecosystems is a relatively new pursuit. This is the first attempt at a coarse-scale freshwater ecosystem classification in British Columbia. For classification purposes coarse-scale freshwater systems are defined as networks of streams, lakes, and wetlands that are distinct in geomorphological patterns, tied together by similar environmental processes and gradients, occur in the same part of the drainage network, and form a distinguishable drainage unit on a hydrography map. Coarse-scale freshwater systems are spatially nested within major river drainages and ecological drainage units (EDUs), and are spatially represented as watershed units (specifically BC Watershed Atlas third order watersheds).

The types and distributions of freshwater systems are characterized based on abiotic factors that have been shown to influence the distribution of species and the spatial extent of freshwater communities. This method aims to capture the range of variability of freshwater system types by characterizing different combinations of physical habitat and environmental regimes that potentially result in unique freshwater communities. An advantage of this approach is that data on physical and geographic features (hydrography, land use and soil types, roads and dams,



Page 9

topographic relief, precipitation, etc.), which influence the formation and current condition of freshwater ecosystems, are widely and consistently available.

Our freshwater ecosystem classification framework classifies environmental features of freshwater landscapes at two spatial scales: ecological drainage units that take into account regional zoogeography, climatic, and physiographic patterns; and mesoscale units that take into account dominant environmental and ecological processes occurring within a watershed. Seven abiotic variables were used to delineate coarse-scale freshwater system types: drainage area, underlying biogeoclimatic zone and geology, stream gradient, dominant lake/wetland features, glacial connectivity, and coastal connectivity. Within each drainage area class (headwaters/small coastal rivers, small rivers, intermediate rivers, large rivers), every watershed was classified according to the dominant biogeoclimatic zone it fell within, its dominant underlying geology, and its dominant stream gradient class. Each of these coarse scale freshwater system types were then further subdivided based on their characteristics of being glacially and/or coastally connected, and if dominant lake and wetland features were present.

Skeena, Nass, Haida Gwaii, and North and Central Coast EDUs collectively consist of 4,476 coarse-scale freshwater ecosystems. These freshwater ecosystems were classified as follows:

• Nass EDU: 755 freshwater ecosystems classified into 74 system types;

• Skeena EDU: 1,333 freshwater ecosystems classified into 182 system types;

• North Coast: 1,773 freshwater ecosystems classified into 150 system types;

• Central Coast: 496 freshwater ecosystems classified into 68 system types; and

• Haida Gwaii: 125 freshwater ecosystems classified into 30 system types.

Freshwater Focal Species Tailed Frog

The tailed frog is a highly localized, specialized species that lives in cool, swift, permanently flowing headwater mountain streams composed of cobble and anchored boulders that provide refuge for tadpoles and adults. In BC, the only Canadian province where it occurs, the tailed frog is found along the Coast Mountains, from the Lower Mainland to Portland Canal, north of Prince Rupert. Although its range in BC is extensive, there are concerns about the status of the tailed frog due to its low reproductive rate, its highly specialized habitat requirements, human activities within its range, and lack of knowledge about minimum viable population size, particularly in fragmented landscapes. Adult tailed frog abundance is positively correlated with the percentage of old-growth forest in a watershed (Dupuis and Friele 2002; Stoddard 2002; Welsh and Lind 2002), most likely because these forests dampen microclimatic extremes (Chen et al. 1993, 1995; Brosofske et al. 1997; Johnston and Frid 2002).

Tailed frog populations in BC have been poorly surveyed. This is the first attempt to model habitat for this species at a landscape scale. Five biophysical conditions were identified as being important for tailed frog habitat: 1) basin area between 0.3 and 10 km2; 2) basins where the bottom elevation < 600 m and the ratio (top elevation – 900)/(900 – bottom elevation) = between 0.0 and 2.0; 3) Watershed ruggedness between 31 and 90%; 4) northerly aspect in Coast and



Page 10

Mountains region and Southerly aspect in Interior regions; and 5) forest cover Age Class ≥ 6. These conditions were spatially modelled within the CIT study area. Habitat areas meeting all five biophysical conditions stated above were classified as being optimal habitat areas for tailed frog. Habitat areas meeting biophysical criteria one to three were classified as being suitable habitat areas for tailed frog.

In total, 4,466 km of suitable tailed frog stream habitat was identified within the CIT study area consisting of 5,155 habitat areas (Map 13). Of this suitable habitat, 2,323 km of stream habitat consisting of 2,486 habitat areas were determined to be optimal habitat that meets all five biophysical parameters within the model. There was a 60% correlation between field survey data and modelled suitable habitat for the tailed frog.

Pacific Salmon and Steelhead

Six species of native anadromous salmon occur in BC: chinook, chum, coho, pink, and sockeye, plus steelhead. They are wide-ranging, migratory species with life histories that integrate marine, freshwater, and terrestrial ecosystems. They are considered a key set of focal species not only because of their highly specialized life histories but also because they play a critical role in the integrity of BC’s coastal ecosystems. They face threats across all life history stages and habitats.

The federal Department of Fisheries and Oceans (DFO) salmon escapement database was used to assess trends in salmon escapement over time and compare biomass estimates across runs. DFO enumerates salmon escapement (number of salmon returning to their natal streams to spawn) annually. The current database contains escapement estimates for Pacific salmon from the 1950s to the present. Although the database has limitations, it contains the best available survey information.

Runs with less than ten years of data were removed from the analysis and labeled as having insufficient data. Trends in escapement over time were calculated for each run. Runs were classified as exhibiting either a stable or declining escapement trend. Runs with slopes ≥ 0.0 were characterized as having stable escapement trends over time. Runs with slopes < 0.0 were characterized as having declining escapement trends over time. Biomass estimates for each run were calculated using the median of ten-year running averages for each run.

Steelhead populations are provincially managed using a different management framework from DFO. Population trends were estimated from catch information collected annually from a mail-out survey, from test fisheries in commercial fisheries with a steelhead bycatch, and from swim surveys, weir counts, and creel surveys. Regional biologists using various combinations of the above methods assigned population status. Results were expressed as the number of spawners divided by the estimated carrying capacity of the river. Populations were assigned to one of three categories based on how they compared to estimated carrying capacity. Salmon population units (SPUs) were derived for each of the species based on existing Fisheries Information Summary System (FISS) point source data, DFO escapement data, and expert knowledge. Escapement trends and biomass estimates were summarized for each SPU (Maps 14 through 21). Biomass estimates were summed for runs within an SPU.

All Pacific salmon species showed a decline in escapement over time with 40% of these declines occurring in high biomass runs. Pink (even-year) runs were found to be in the best condition and chum in the worst condition. Steelhead were found to be in stable condition across all EDUs except for the Central Coast. The Central Coast had 80% of its SPUs exhibiting declining trends.



Page 11

Haida Gwaii had 76% of its SPUs in decline. coho appeared the most stable with 53% of SPUs containing stable runs. All summer and winter steelhead SPUs were stable. Most species’ SPUs were of low biomass (58%) except for sockeye in which 83% of SPUs were high biomass runs. Ninety-seven percent of summer steelhead SPUs were in decline compared to only 7% of winter steelhead SPUs. The majority of pink (even and odd year) and chum SPUs had high biomass runs. The majority (76%) of SPUs were in decline within the Haida Gwaii EDU. The Central Coast EDU was found to be in the worst condition of all EDUs with 80% of its SPUs and 48% of its steelhead (summer and winter) SPUs in decline. All salmon species were found to have greater than two-thirds of their SPUs in decline except for even-year pink (53% in decline) and steelhead (33% winter and 12% summer in decline). Chum had the highest proportion of SPUs in decline (77%) followed by coho, sockeye, chinook, and pink (odd-year).

These results indicate downward trends in escapement over time for all five species of Pacific salmon throughout the entire CIT study area as well as prominent declines in winter and summer steelhead within the Central Coast. Causes of decline stem from multiple factors including changes in ocean regime patterns, marine survival conditions, fisheries management, land use, hatcheries, and over-fishing.

Nearshore Marine Targets

Introduction

We focused on the intertidal and nearshore zones in this analysis. We identified 72 conservation targets of which 59 were ecosystem targets (shoreline ecosystems) and 13 were special element targets (occurrences of marine species). This set of targets was selected to represent the full nearshore marine biodiversity of the region and to highlight elements that are threatened or declining (i.e., seabird species), or that serve as good indicators of the health of the larger ecosystem (i.e., intertidal habitats). The waters of the ecoregion were stratified into 8 sections based on British Columbia’s marine ecosections (Map 22). Ecosections are characterized as unique physiographic, oceanographic, and biological assemblages that are related to water depth and habitat (pelagic versus benthic). We performed raster-based analyses to develop a human impact score for all shoreline planning units (Map 23 and 23a). We also used a stratification scheme based on project regions (Map 24).

Nearshore Marine Special Elements

We included as special elements species that are imperiled or keystone species and are not likely to be protected by ecosystem representation. With these criteria we included eight marine species as targets. These forage fish and seabird targets are represented as two critical life stages, spawning aggregations and breeding colonies.

We had comprehensive coverage for herring and seabird colonies. The latter data set includes the locations of all known seabird colonies along the coast of British Columbia. Fifteen species of seabirds, (including two storm petrels, three cormorants, one gull, and nine alcids) and one shorebird (Black Oystercatcher) breed on the coast of British Columbia. We selected those species that are considered imperiled or vulnerable in British Columbia (S1–S3 status). There were seven species of seabirds selected to represent colonies along the coast of British Columbia: Thick-billed



Page 12

and Common Murre, Cassin’s Auklet, Ancient Murrelet, Horned and Tufted Puffin, and Brandt’s Cormorant. These data are represented as point locations with attributes describing their location. We used location descriptions to assign the colonies to specific shoreline segments.

Our biggest data gap concerned invertebrates. The Conservation Data Centre (CDC) does collect element occurrence data for some invertebrate species, but the data are not comprehensive throughout the study area. Without a comprehensive, continuous survey effort we were limited by the places where species were found and therefore did not have a sense of abundance across the region.

Nearshore Marine Ecosystem Representation

The ecosystem targets comprise shoreline ecosystem types derived from the shorezone mapping system. British Columbia’s shorezone mapping system is based on shore types, which are biophysical types that describe the substrate, exposure, and vegetation across the tidal elevation, as well as the anthropogenic features. The BC shorezone mapping system is built on shore types that aggregate precise community or habitat types according to their landform, substrate, and slope. There are 34 coastal classes and 17 representative types within the classification system. Wave energy classes were aggregated to five classes, which when multiplied by the 17 representative types yielded as many as 85 classes. In actuality there are only 63 shoreline categories available after compiling all six project regions of shorezone. Four categories, three “man-made” types and one “unidentified,” were not considered ecosystem targets, bringing the total number of targets down to 59.

The shorezone mapping system also identified intertidal vegetation and habitats. “Bio-bands” are assemblages of intertidal biota, visible from the air and named for dominant species or species assemblages. With these bio-bands we identified the intertidal range biota in the nearshore, including saltmarsh, eelgrass, surfgrass, and kelp. These vegetation types form the major habitats of the nearshore zone and are the best surrogates at this scale to represent a range of habitats. We also used the “habitat observed” category from the shorezone data set to capture the most diverse part of the intertidal zone.

Although we included “estuaries” as targets in the analysis, we did not have complete information on all estuaries in B.C. Both the terrestrial ecological systems and the shoreline ecosystem targets identified saltmarsh communities. In both cases the definition of estuaries was associated with a vegetation type. Other estuaries were captured in the shoreline ecosystem categories as “mud flat,” “sand flat,” and “sand and gravel flat.” The shorezone data do not fully represent the spatial delineation of estuaries as polygons. Many estuaries are left out of the data sets because they were not identified as the dominant feature in the intertidal.

Offshore Marine Targets

Special Elements and Focal Species

The CIT offshore marine ecosystem spatial analysis includes 93 features, both biological and physical. We considered the following focal vegetation species: eelgrass, kelp, marsh grasses, surf grasses, and a general shoreline vegetation class, aggregated from the BC Shorezone classification. All major BC breeding seabird populations and colonies were considered (Map 25):



Page 13

Ancient Murrelet, Black Oystercatcher, Cassin’s Auklet, Cormorant species, Glaucous-winged Gull, Pigeon Guillemot, Puffin species, Rhinoceros Auklet, and Storm Petrel species. In addition, very small islets, far from shore were also considered as surrogates for unsurveyed colonies. Seabirds are known to prefer certain marine waters. These we treated as “habitat capability” layers. We considered pelagic seabirds, waterfowl loons, and shorebirds. Moulting sea ducks (Scoters and Harlequin Ducks) inhabit certain nearshore BC waters during summer (Map 26). These areas were also considered separately for each species grouping.

Anadromous streams were captured using a species richness x stream magnitude ranking (Map 27). Eight of BC’s nine anadromous species were considered (eulachon, the ninth, was treated separately). About one out of 10 BC stream systems were considered likely to support significant numbers of anadromous species. Steller sea lion haul-outs and rookeries were ranked on a scale of 1–4 based on population density. Herring spawn shorelines were ranked on a density measure based on DFO’s Spawn Habitat Index, using the latest available times series data.

We considered five special elements, on account of their rare or threatened status: hexactinellid sponge reefs, eulachon estuaries, sea otter, estuaries containing red or blue listed species, and Marbled Murrelet marine habitat. We also considered areas known to harbour large habitat-forming corals, which may well be threatened or endangered, but due to a lack of surveys their status remains largely unknown.

Offshore Marine Ecosystem Representation

We included two separate indicators of distinctive habitats for ecosystem representation in the offshore marine realm: benthic topographical complexity and high current. For each of these analyses, the study area was stratified into four ecological regions—inlets, passages, shelf, and slope—based on available substrate and depth information (Maps 28 and 29). Areas of high taxonomic richness are often associated with areas of varying habitat. Complex habitats may also exhibit greater ecosystem resilience and resistance to invasive species. Benthic topographical complexity is indicated by how often the slope of the sea bottom changes in a given area. Benthic complexity considers how convoluted the bottom is, not how steep or how rough, though these both play a role. A measure of benthic complexity will often identify physical features such as sills, ledges, and other habitats that are associated as biological “hotspots” providing upwellings, mixing, and refugia. We examined benthic complexity separately within each of the four Ecological Regions (Map 30).

The high current layer was extracted from the BC Marine Ecological Classification, as well as incorporating additional local knowledge. High Current is defined as waters that regularly contain surface currents (tidal flow) greater than 3 knots (5.5 km/h or 1.5 m/s). These are areas of known mixing and distinctive species assemblages. In addition, high current areas often represent physical “bottlenecks” to water movement and as such are important to larval transfer and nutrient exchange. The constant re-suspension of nutrients in particular is most likely responsible for the rich biota of the south Central Coast passages. Annual primary productivity in tidally mixed areas tends to be above average for coastal waters. High current areas were considered separately for each of the four Ecological Regions (Map 30).



Page 14

Setting Goals

Explicit and quantitative goals are fundamental to systematic conservation planning. Establishing goals is among the most difficult—and most important—scientific tasks in conservation planning (e.g., How much protected area is enough? How many discrete populations and in what spatial distribution are needed for long-term viability?). Goals for conservation targets specify the number and spatial distribution of occurrences. A broad goal is to conserve multiple examples of each target, stratified across the region in such a way that the variability of the target and its environment is captured in the site selection. Replication of occurrences of each target must be sufficient to ensure persistence in the face of environmental stochasticity and the likely effects of climate change. Strategies for focal species emphasize persistence of populations or metapopulations, whereas ecosystem-level conservation (e.g., representation) often invokes operation of ecological processes and maintenance of ecological integrity.

As species richness and endemism increase, so does the amount of area needed to represent all elements. Because northern temperate regions have lower biodiversity and fewer narrow endemics than found in the tropics, it is expected that less protected area is necessary to represent each biodiversity element at least once. Importantly, site selection algorithms, by themselves, do not address the more difficult and real-world questions concerning the area needed to maintain viable populations of species (i.e., as opposed to single occurrences) and overall ecological integrity. We did not have the time or funding to gather data and perform detailed, spatially explicit population modeling for the selected focal species. Therefore, we were unable to evaluate the potential population viability of these species in alternative networks of reserves compared to the current network. Nonetheless, previous regional scale biodiversity assessments have identified representation benchmarks that fit within the 30 and 70% representation thresholds cited by the CIT’s EBM Framework and that generally require between 25 and 75% protection of land area. Higher proportions of the landscape devoted to protection could certainly be argued for, particularly given the sparse data on the ecological dynamics and requirements to maintain focal species and natural processes. Indeed, the precautionary principle dictates higher levels of protection to buffer against these uncertainties.

For the CIT ESA, initial goals were set for all of the targets based on their geographic scale, distribution, and spatial pattern. A GIS data set was then created for input into the site selection algorithm. Given the time and resources available, we determined that the best approach to setting goals draws on the EBM framework by setting a range of goals that can be used to construct separate portfolios for several goal levels within that range. Using this approach, a series of potential conservation solutions were created for 30, 40, 50, 60, and 70% goal settings, wherein these percentage goals were applied uniformly across all ecosystem and focal species targets. These solutions were then used for purposes of prioritization, allowing the team to compare areas that were necessary to satisfy all goal scenarios including the minimum goal set, to areas only selected in larger goal sets, and to those areas never selected, regardless of the goal setting.

Threats (Human Impacts) Analysis

We developed spatial tools to summarize relative levels of human impacts, using the watershed units as our primary analytical units. Watersheds form a logical unit to summarize relative levels



Page 15

of human impacts because ecological linkages within watersheds tend to be stronger than linkages between watersheds.

We applied a classification based on Moore (1991), with some modifications (Map 31). Human impacted area was calculated by combining human altered area (clearcut, urban, agriculture) with a 200 m buffer area around roads. Overlapping areas were treated as impacted. We calculated impacted area as a percentage of potential vegetated area, which was calculated as a sum of natural vegetated area, human altered vegetated area, and urban area. Watersheds with more than 2% of their area affected may still be considered as ecologically intact, depending on both the cumulative impact of human alteration and the spatial location of human alterations. To identify such watersheds, we used two additional factors for assessing the overall impact, 1) proximity of impacts to rivers and streams, and 2) road density. Classification thresholds for modified areas were set at levels where road density has been demonstrated to impact grizzly bear populations.

Although Moore (1991) restricted analysis to watersheds greater than 5,000 ha, we also sought to identify relatively intact watersheds at multiple spatial scales. Small intact watersheds may be sufficient for harboring viable occurrences of some species and/or community types (e.g., rare plant communities), but larger, contiguous intact areas are necessary to conserve viable populations of vertebrates. Because of their value for salmon populations and global rarity, entire river systems that are relatively intact represent key areas for conservation.

Spatial Analysis

For the CIT ESA, the challenge is to take an analysis of special elements, ecosystem representation, and focal species, and create a spatially explicit assessment of where regional biodiversity values are located and what condition they are in. To this end, we employed computerized site selection algorithms in combination with the impacts analysis described above in a GIS environment. Importantly, site selection algorithms, by themselves, do not address the more difficult and real-world questions concerning the area needed to maintain viable populations of species and the persistence of biodiversity. The analysis can however provide stakeholders and decision makers with reference points for exploring how potential conservation solutions might be distributed across the study area. This information can help direct where conservation values are highest and which geographies might best be suited for no or low risk landscape management practices. Conversely, this information may be useful in helping to identify locations where resource development might occur with minimal impacts to ecological values.

Terrestrial and Freshwater Analysis

For the terrestrial and freshwater analysis, we used the site selection software SITES. SITES applies an algorithm called “simulated annealing with iterative improvement” as a method for efficiently selecting sets of areas to meet conservation goals. The algorithm attempts to minimize portfolio “cost” while maximizing attainment of conservation goals in a compact set of sites. This set of objectives constitutes the “Objective Cost function:” Cost = Area + Species Penalty + Boundary Length, where Cost is the objective (to be minimized), Area is the number of hectares in all planning units selected for the portfolio, Species Penalty is a cost imposed for failing to meet goals, and Boundary Length is a cost determined by the total boundary length of the



Page 16

portfolio. SITES attempts to minimize total portfolio cost by selecting the fewest planning units and smallest overall area needed to meet as many goals as possible, and by selecting planning units that are clustered together rather than dispersed.

We used 500 ha hexagons as planning units for the SITES analysis. Use of uniform-sized planning units avoids the area-related bias that can occur when differently sized planning units, such as watersheds, are used. We applied four different protection scenarios: no protected areas “locked in” the outcome; existing protected areas “locked in”; existing plus candidate protected areas “locked in”; and existing plus candidate plus option areas “locked in.” Candidate and option areas were only available for the Central Coast region.

All else being equal, planning units with lower levels of human impact should be chosen over those with higher levels of impact, to select areas in better condition with higher chances of viability. Thus, rather than simply using the number of hectares in each planning unit for the Area component of our SITES analyses, we developed a Suitability Index, i.e., a Cost Index based on the human impact data. Human impacted area was calculated as described above. To account for planning units with relatively little vegetated productive areas (and consequently little developable area and little productive habitat) we used the following suitability index: Cost Index = Planning Unit Area + Planning Unit Area * Human Impacted Area/Potential Vegetated Area. Potential Vegetated Area was calculated as the sum of vegetated habitat plus the sum of clearcut and urban areas. Because the SITES algorithm seeks to minimize total portfolio cost, it selects planning units with low cost (i.e., low impact) over high-cost areas whenever possible. We used boundary length modifiers (BLM) of 0.001, 0.01, and 0.1 to include a range of planning unit clustering in our final combined runs. This range allows us to explore issues of several small clusters versus fewer large clusters.

We made 20 repeat runs (each comprised of 1 million iterations of planning unit selection) for each of 15 combinations of boundary length modifier (three levels) and goal (five levels) for each of the four protected area scenarios. Thus, for each protection scenario we used a sum of 300 SITES runs that resulted from 300 million iterations of the simulated annealing algorithm. Hexagons chosen frequently represent places more necessary or irreplaceable for biodiversity conservation, while those chosen few times represent locations where similar biodiversity is found elsewhere or where human impacts are significant. We used five different goal levels: 30, 40, 50, 60, and 70%, as described above.

Nearshore Marine Analysis

For the nearshore marine analysis we used a site selection algorithm called MARXAN, which employs the same basic simulated annealing procedure as the SITES algorithm. Shoreline ecosystems were analyzed as linear segments. These units vary widely in length and extend over the entire coastline. Forage fish spawning sites and seabird colony data were attributed to the shoreline segments to represent the nearshore zone.

We included in the analysis a Suitability Index, or “Cost Index,” which tends to reduce representation in places with high human impacts. Impacted areas were calculated as described above, but three cost parameters were added: aquaculture tenures, enhancement facilities, and hatcheries. We performed raster-based analyses to develop a human impact score for all shoreline planning units (Map 23). To build the Cost Index, we took the mean planning unit length and the



Page 17

sum length of human impacts within that planning unit: Cost = Mean Planning Unit Length + (Mean Planning Unit Length * Human Impact Score). We used the mean value of all shoreline planning units instead of calculating cost equal to length.

We developed a linear boundary modifier that clumped adjacent linear segments along the shoreline. The algorithm was therefore able to assemble small fragments of shoreline into more continuous stretches (i.e., select an entire islands’ shoreline). We set the boundary length modifier to 0.1 for all MARXAN scenarios. We set penalty factors based on the importance of the target; a high penalty factor is assigned to a high priority target.

Our approach for building a nearshore marine portfolio combined expert input (through a technical workshop) with spatial analysis. One of the objectives from the technical workshop was to select biologically diverse coastal sites based on expert knowledge. These initial “seascape sites” were chosen to capture relatively large, intact ecosystems that represent the region’s nearshore biodiversity. Experts were asked to select areas along the B.C. coast using three criteria: 1) Large nearshore sites that are important for marine biodiversity. 2) Sites must be stratified across the ecoregion. 3) There should be diversity in the types of sites chosen, i.e., good for birds, good for invertebrates, etc. The results were the identification of 18 seascape sites, which helped guide subsequent analysis.

We evaluated 12 different MARXAN scenarios to test the irreplaceability and sensitivity of the site selection. Irreplaceability analyses indicate which sites are consistently chosen. Planning units that get chosen the most often are the least replaceable. With each of the 12 scenarios run 20 times, we had a gradient from 0 to 240 solutions. Within each scenario the algorithm did 1,000,000 iterative selections per run. Two multiple goal scenarios were run: one optimizing for 10% of the entire shoreline and another for 20%. We used three stratification schemes to divide the nearshore zone: marine ecosections, project regions, and no stratification.

Offshore Marine Analysis

The offshore marine analysis applied 500 ha hexagons as planning units, as in the terrestrial/freshwater analysis, and the MARXAN site selection algorithm, as in the nearshore marine analysis.

Options and Scenarios

To facilitate examination of spatially explicit conservation solutions, we made use of SITES and MARXAN to create a series of potential conservation solutions that in combination, are referred to as Options and Scenarios. At the heart of this exercise was an attempt to prioritize solution outputs according to criteria related to conservation utility, which was based on target information and SITES/MARXAN outputs, as well as conservation condition, based on the evaluation of human impacts within the study area.

For the CIT ESA we used SITES and MARXAN summed solution scores (i.e., showing the number of times individual planning units were selected during runs of the model), as a broad measure of conservation value. For the terrestrial/freshwater analysis, a conservation utility score was derived for analysis units based on the frequency by which any one planning unit was



Page 18

selected in the total of 300 SITES runs (20 repeat runs x 3 boundary length modifiers x 5 goal settings), that were performed.

Scores for both planning units were then grouped into five classes based on the quintile scores of the summed solutions. A planning unit selected 180 or more times in SITES fell into the top two quintiles, or top 40%, of the solution and was scored as having high conservation utility. Analysis units with scores in the middle quintile were scored as medium conservation utility, and those in the 4th quintile were scored as low conservation utility. Those units with a score in the lowest quintile were not ranked.

Condition was used as a surrogate measure for target viability in the CIT ESA and was evaluated using the human impacts information described above. The six impact classes were simplified into the three broad condition classes--intact, modified, and developed.

Alternative Options and Scenarios for Conservation

To explore the interaction between conservation utility and condition, analysis units were clustered into three conservation tiers based on a conservation utility and condition matrix. Under this framework, areas ranked as intact or modified that also hold high conservation utility, or intact areas with medium conservation utility, were ranked as Tier 1. The middle tier (Tier 2) represents those areas with high utility but which are highly impacted, or areas with low utility, but which are intact, or areas that fall within the mid-range of both criteria (medium utility/modified condition class). Tier 3 represents those analysis units that are developed and which have a medium or low conservation utility. These three tiers constitute three conservation options that can be evaluated against various land use scenarios.

The Central Coast LRMP tables have already proposed several potential land use scenarios, and we wanted to evaluate each in terms of their performance against the three conservation options being generated by the ESA. To facilitate this comparison, the protected areas described by each scenario were locked into the SITES solution for terrestrial/freshwater runs. Four alternative land use scenarios were evaluated as follows:

1. Unconstrained Analysis.

2. Base Case – existing protected areas locked into SITES.

3. Candidate Case – existing protected and CCLRMP Candidate Areas locked into SITES.

4. Option Areas Case – existing protected areas, CCLRMP Candidates, and Option Areas locked into SITES.

To compare between and within options and scenarios, potential solution sets from summed runs were evaluated against three goal thresholds: 30, 50, and 70%. For each scenario, performance of conservation tiers (options) was assessed both in terms of effectiveness, as measured by the proportion of targets that met or exceeded the goal threshold, and efficiency, the proportion of the study area required to meet the threshold.



Page 19

Relative Biodiversity Index

To further support decision-making throughout the LRMP process, stakeholders also requested more transparency in the presentation of data on conservation values. Specifically, LRMP table members were interested in maps that showed biodiversity values broken out by system type and species, and asked that these values be displayed on a continuum, using third order watersheds as a reporting unit. To meet these needs, the ESA team created a second index of conservation value based on the existing data inputs that were used in SITES runs, and applied it to third order watersheds. This relative biodiversity index (RBI) required calculating for each ESA target the relative contribution of each watershed to the total available habitat or ecosystem extent within each LRMP boundary.

These scores were calculated independently for each focal species and for each ecosystem type and also translated into a total relative biodiversity index (RBI) for all third order watersheds in each LRMP area. Impacts are not taken into account for this measure.

Spatial Analysis Results

Terrestrial/Freshwater Spatial Analysis Results

Summed run solutions for each of the four land use scenarios are displayed in Maps 32 through 35, with Tier 1 and 2 analysis units highlighted in Maps 36 thorough 39. While the pattern of conservation utility differs between each land use scenario, there are only small differences among the overall performance of the solutions. This similarity is true in regard to both the solution’s efficiency, measured by the amount of area swept into each conservation option or tier, and effectiveness, measured by the conservation goals reached by those tiers.

Depending on the conservation scenario, between 44 and 50% of the highest conservation utility planning units (equal to 44–50% of the landbase of the study area), were required to satisfy the 30% goal threshold for targets. Increases in solution area resulted in approximately proportional increases in solution effectiveness. To achieve 50% goals for most conservation targets, as much as 60–70% of the study area was required. After approximately 50–60% of the study area (50–60% of the highest utility analysis units), has been incorporated into the solution, subsequent improvements in meeting goals required proportionally larger areas of land and water. That is, the incremental increase in goal achievement declined. This effect is much more pronounced when the 70% goal threshold is examined. In this case, inefficiencies arise to the point where the inclusion of more planning units yields only minute improvements relative to meeting goals. It is important to note that using the SITES summed solution scores for identifying core conservation areas, does not represent an “optimal” design methodology. So called optimal SITES designs are created using the “best run” feature of SITES, but given the considerable uncertainty around conservation goals, boundary length modifiers, and the direction of LRMP negotiations, the team elected to use a summed run approach so as to convey a range of potential conservation solutions as opposed to a single “correct” solution. It is recommended that as agreements around protected areas are sequentially reached, that SITES optimal runs be used as a means of evaluating the success of proposed reserves in meeting representation benchmarks, and for identifying remaining areas needed for conservation. If protected areas are well chosen based on high conservation values, these optimized runs should yield more efficient solutions, i.e., the



Page 20

percentage of area captured in the SITES run should be closer to the percentage of total ESA targets meeting representation goals.

Under any combination of proposed and existing protected areas we see that less than 20% of the conservation elements being targeted by the CIT ESA meet a 30% goal threshold. However, it is apparent that there are considerable conservation values within the areas that have been designated and proposed. Up to a quarter of the region’s Tier 1 analysis units would be captured in a scenario that designated Candidate and Option Areas as protected.

Summaries of conservation tiers for each land use scenario are presented in Maps 37 to 39. Tier thresholds for conservation utility were based according to quintiles of the mean summed solution scores of the analysis units within each watershed.

Results of RBI scores for third order watersheds are displayed in Map 40. High total RBI scores may occur as a result of many co-occurring conservation targets in a watershed, but may also result when a watershed contains most of the extent (or available habitat value) for a single target. This latter instance occurs when a watershed contains examples of targets that are more spatially limited within the study area. This methodology also favours larger watersheds over smaller since it is a cumulative score and the more area available, the higher the likelihood that more ecosystems and habitat will be present. Since condition of the watershed is not incorporated into the measure as currently displayed, some high value, but developed watersheds are ranked higher than more pristine, intact watersheds. In the future, this score should be adjusted by the watershed’s condition class to account for impacts.

CC LRMP Representation Analysis

The ESA team had the opportunity to perform a “gap” analysis on a number of proposed protected areas solutions emerging from the CC LRMP table in the fall of 2003. Results from the analysis of the CC LRMP final agreement of December 12, 2003, are presented in Appendix 5.1 and a specific example, the analysis of focal species habitat representation, is presented in Figure 6.4. Previous to the December 12 agreement, existing protected areas in the region had represented 10% or less of ESA focal species habitat. However, the combination of Candidate Areas and Option Areas agreed to through the LRMP process increased representation of habitat by roughly three to five-fold, depending on the species. Substantial and important gains were also achieved for focal ecosystems, old-growth ecosystems, and freshwater systems (Appendix 5.2). A closer examination of focal species habitat differentiated by ecosection does however demonstrate significant shortfalls in representation of some ecosections of the CC LRMP area—most notably in the Outer Fjordlands ecosection. Similar shortcomings exist for focal ecosystem targets and old growth in this ecosection, and to a lesser extent, in the Northern Pacific Ranges ecosection, Nonetheless, the authors are unaware of other North American examples where a protected areas system has so specifically captured such a significant proportion of a region’s biodiversity values.

Nearshore Marine Spatial Analysis Results

Sensitivity analyses were conducted by selecting thresholds in the summed solution gradient and evaluating how well nearshore targets were captured in solutions. The area required to meet goals changed as we lowered the threshold along the summed solution gradient. We set three



Page 21

thresholds to illustrate this: 10% (Option A), 20% (Option B), and 30% (Option C) of the entire shoreline length. The summed solution gradient from the 12 MARXAN scenarios provided the basis for setting these thresholds. Several options for displaying and interpreting these results using shoreline units are displayed in Maps 41–44.

Offshore Marine Spatial Analysis Results

Conservation utility was ranked according to the number of times each planning unit was selected in 2,400 MARXAN solutions (displayed in Map 45). The examination of various clumping values indicates that regardless of whether reserves are many and small, or few and large, certain areas recur over the course of many runs. For example, within the Central Coast, the larger areas of high conservation value that emerge include hexactinellid sponge reefs, Goose Islands, Bardswell Islands and vicinity, Rivers Inlet, Scott Islands, Entrance to Queen Charlotte Strait, Broughton Archipelago, Head of Knight Inlet, and Cordero Channel.

Although these areas alone would not constitute a fully representative Central Coast conservation portfolio, it is very likely that were they not included, such a portfolio would be difficult or impossible to achieve. Thus, regardless of the exact percentages chosen by planning processes, and the exact shape of the boundaries, we would expect the bright yellow areas of Map 45 to be key components of most conservation planning. Larger areas of high conservation value within the North Coast include hexactinellid sponge reefs, West Aristazabal Island (and NW Price I.), Kitimat Arm, Anger Island and vicinity, SW and N Porcher Island, Kitkatla Inlet, S. Chatham Sound, and Mouth of the Nass River.

Larger areas of high conservation utility within the Haida Gwaii waters include W. Dixon Entrance, Naden Harbour, Masset Inlet, Skidegate Inlet (Kagan Bay), and South Moresby Island.

Areas of high conservation utility alone would not constitute a fully representative conservation portfolio. The individual network solutions produced by MARXAN can be diverse. Such diversity allows for flexibility when considering external factors, such as user interests, parks, local politics, access, and enforcement.

Conclusions

The CIT ESA constitutes the first regional biodiversity analysis of its kind for BC’s Central Coast, North Coast, and Haida Gwaii. Our purpose was to assess conservation values for terrestrial, freshwater, and marine ecosystems and species, to prioritize the region for conservation attention, and to provide decision-makers with an assessment tool that evaluates both current and proposed protected areas scenarios. The models for focal species and ecosystems must be recognized as regional in scale, and the information is not appropriate, or intended for, decision-making at stand or operational scales. Further, limitations in time, human resources and financial resources required the ESA to limit its analysis to a static assessment of conservation values, as they currently exist on the landscape. While much debate continues regarding thresholds for sufficiency for species and ecosystems (i.e., how much is enough?), the ESA clearly demonstrates that most ecosystem types native to the study area are not currently represented in existing protected areas. The ESA should be used to identify areas of highest conservation priority and ideally, protected areas will be situated in



Page 22

locations overlapping with Tier 1 and 2 conservation areas identified by this report. On a more practical level, the ESA team recognizes that negotiations around protected areas involve a sequential, step-wise approach in which trade-offs are made between competing ecological, cultural, and economic interests. Therefore, the ESA has also provided maps and data sets describing a Relative Biodiversity Index, which specifically addresses the needs of stakeholders in such a decision-making environment.

Once protected areas have been agreed upon by planning tables, efforts must focus on using the ESA to assess how well species and ecosystems are represented, and information emerging from gap analysis for all three LRMP areas needs to be highlighted for future work around ecosystem-based management. Examples of habitats and systems not well represented need to be identified, and well protected through additional landscape, watershed, and stand-level reserves and prescriptive management guidelines. Next steps should include analysis of connectivity between proposed protected areas, and an examination of their size, configuration and management. Such work is critical for exploring questions related to the adequacy of proposed protected areas for ensuring the long-term viability of the species and systems represented within.

According to the CIT’s Ecosystem-Based Management Planning Handbook, if the study area is to be managed for low ecological risk, 70% of the range of natural variation of habitats and ecosystems needs to be represented and conserved regionally. EBM provides for a number of ways for such goals to be achieved at various geographic scales. From a regional perspective, one of the most certain ways to carry out such conservation is to institute “no risk” management regimes to capture a significant proportion of that 70% representation goal. The authors of this report acknowledge the uncertainties surrounding recommendations regarding the exact percentage of the study area that needs to be set aside in parks and protected areas. Nonetheless, we cannot escape the conclusion that more protection is better than less, and that more certain protection (e.g., park designation) is better than less certain protection (e.g., results-based regulation).

Substantial work remains to be done with regard to regional analysis of biodiversity priorities and management in coastal BC. The authors strongly urge that improvements be made through the integration of the ESA regional analysis into the proposed ongoing analysis to be undertaken by the coastal Ecosystem-Based Management Council that is being negotiated as part of the LRMP process. Future efforts must be directed at analyzing the outcomes of LRMP and LUP negotiations and at bringing new research and data into the analysis, including robust viability assessments and the addition of temporal modelling of dynamic ecosystem processes at the regional scale. Much work also remains to be done in terms of integrating marine and terrestrial analysis. Further, the regional scale analysis described in this report must be fully integrated into an EBM framework that accounts for conservation at finer scales—and vise versa—stand and operational management plans must be rolled-up and assessed across the region to provide a full assessment of the cumulative effect of resource management decisions.



Page 23

1.0 INTRODUCTION

This report presents the methods and results of a comprehensive ecosystem spatial analysis for the Haida Gwaii, Central Coast, and North Coast regions of British Columbia. This study assesses conservation values for terrestrial, freshwater, and marine ecosystems across this vast region. The methodology employed in this study integrates geographic information system (GIS) analysis of rare species and other unique features (“special elements”); representation of terrestrial, freshwater, and marine ecosystem units; and conservation of suitable habitats for species of special interest (“focal species”).

Conservation planning on a regional scale has become the standard approach for organizations and agencies worldwide interested in the conservation of biodiversity and other ecological values. Whereas much of recent conservation history in North America has been dominated by actions often described as “piecemeal,” “species-by-species,” or “site-by-site,” in large part stimulated by the requirements of species-based legislation (Noss et al. 1997), conservationists today focus increasingly on ecosystems, landscapes, and ecoregions. Species conservation is no longer piecemeal, as the conservation of individual species is now interpreted within the broader context of maintaining the structure, function, and composition of ecosystems within a natural or historic range of variability (Franklin et al. 1981; Landres et al. 1999; Swetnam et al. 1999). Moreover, modern conservation seeks to conserve species and other elements of biodiversity before they become threatened or endangered and before conflicts between conservation and economic activities escalate out of control.

Regional conservation planning differs from conventional land use planning in that regions are defined ecologically rather than politically (Noss and Cooperrider 1994). Many conservation agencies and non-profit groups, in North America and elsewhere, base their planning on the boundaries of ecoregions—large areas distinguished by similarities in climate, landform, soils, vegetation, and natural processes (Dinerstein et al. 1995; Bailey 1998; Ricketts et al. 1999; Groves 2003). Such ecoregions regularly overlap provincial and national boundaries. In 1996, The Nature Conservancy initiated its approach to ecoregional planning in the United States, drawing on experience from ad hoc regional conservation plans nationwide. Shortly thereafter, World Wildlife Fund, working with a number of experts, provided a conservation assessment of the ecoregions of the United States and Canada (Ricketts et al. 1999), one of many assessments the organization has undertaken worldwide. Meanwhile, since 1991, The Wildlands Project and cooperating groups have been developing regional conservation plans and reserve network designs across the United States, Mexico, and Canada, drawing on prototypes developed earlier (e.g., Noss 1987, 1993). Today, The Nature Conservancy and the Nature Conservancy of Canada have completed or are in the process of completing ecoregional plans in over 100 ecoregions in North America and elsewhere in the world

A fundamental requirement of regional conservation planning is that it be systematic. As described by Margules and Pressey (2000), systematic conservation planning is superior in many ways to opportunistic or politically biased planning because it: 1) requires clear choices about the features to be used as surrogates for overall biodiversity, 2) is based on explicit goals, preferably translated into quantitative, operational targets, 3) recognizes the extent to which conservation goals have been met in existing reserves, 4) uses simple, explicit methods for locating and designing new reserves to complement existing ones in achieving goals, 5) applies explicit criteria



Page 24

for implementing conservation action on the ground, and 6) adopts explicit objectives and mechanisms for maintaining the conditions within reserves that are required to foster the persistence of key natural features, together with an effective monitoring and adaptive management program. The CIT ESA is designed to directly address the first four of these objectives. The last two objectives concerning conservation action and management can be addressed only after ESA information is integrated into an ecosystem-based management framework. Finally, regional conservation planning is precautionary. Although reserve selection algorithms, based on mathematical models that emphasize efficiency, attempt to capture the most biodiversity in the least area, the minimal area is properly interpreted as the area sufficient and essential to meet the stated conservation goals and objectives. “Sufficient” implies that the action can be fully expected to attain the stated goals; “essential” implies that, without the action, the goal or objective will not be attained. While actions such as protecting more land than necessary to assure viability of species and ecosystems are ideally avoided, the precautionary principle, which is becoming well accepted in many fields (Peterman 1990; Shrader-Frechette and McCoy 1993; Taylor and Gerrodette 1993; Noss et al. 1997), suggests that, in cases of uncertainty, it is better to risk protecting too much than too little. This precaution can be implemented in conservation planning by setting ambitious goals, while using the best available science to reduce uncertainty over time. The CIT has proposed to address improvement of the scientific understanding of the region and the reduction of uncertainty over time via an adaptive planning and management process applied as an ecosystem-based management framework. It is critical that the ESA, as a regional conservation assessment, be integrated into that framework.

As a coarse-scale, relatively low-resolution assessment, a regional conservation plan such as the ESA is not intended to offer detailed guidance for site-level management of either protected areas or the landscape matrix. Such guidance is better provided by through ecosystem-based management at the site and landscape levels (e.g., the province’s EBM handbook). The ESA, like other regional conservation assessments, takes a macroscopic view of the region, and is useful for 1) highlighting areas of regional biological significance; 2) portraying the spatial pattern of high-value sites on a broad scale; 3) illuminating the landscape context of these sites; 4) assessing the conservation needs of wide-ranging (i.e., “regional-scale” and “coarse-scale”) species; and 5) identifying priorities for further, more detailed and research on finer spatial scales. Again, regional conservation planning is not intended to address site-level issues, which require high resolution of habitat features. Hence, regional-scale planning is not appropriate for developing conservation guidelines for species, such as many small vertebrates, invertebrates, and plants, whose distributions and viability are determined by habitat features and processes operating at fine spatial scales. For a comprehensive assessment of conservation and management needs, regional-scale planning should be followed by progressively more detailed research and planning at landscape, watershed, and local scales. We explicitly identify, in our chapter on Next Steps and Future Work, the need to integrate the results of the ESA with the work of the proposed coastal Ecosystem-Based Management Council that is being negotiated as part of the LRMP and ongoing government-to-government process. We strongly endorse future analyses at sub-regional, landscape, watershed, and stand scales, which will complement the information developed through the ESA study.



Page 25

1.1 The Planning Process

The ESA attempted to integrate three basic approaches to conservation planning: 1) protection of special elements; 2) representation of a broad spectrum of environmental variation; and 3) protection of habitats of focal species. Together, these three tracks, which will be described in more detail below, constitute a comprehensive approach to regional conservation planning (Noss et al. 1999, 2002).

1.2 Study Area and Objectives

1.2.1 Study Area

North and Central Coastal British Columbia (BC) covers the coastal waters, islands, and watersheds of the Canadian Pacific from the Alaskan border south to the Strait of Georgia and from the summits of the coastal ranges west to the continental slope. It includes Haida Gwaii/Queen Charlotte Islands (QCI) and northern Vancouver Island. The region has a land area of 11 million ha (the size of Iceland or Guatemala) but a population of just 90,000—a third of Iceland’s and less than a hundredth of Guatemala’s. Its sea area is another 11 million ha.

The region is the heart of the Northeast Pacific archipelagos coast, one of the world’s three large glaciated leading edge coastal zones (the others being Norway and southern Chile). These zones are on the leading edge or collision margin of tectonic plates and have been sliced and diced by glaciations. Hence they have narrow to nonexistent coastal plains, drop sharply from mountain heights to ocean depths, and are incised by fiords, fissured by channels, and broken into islands large and small.

Biologically and culturally, the Northeast Pacific is the most diverse of the three zones. In particular, North and Central Coastal BC includes the world’s largest tracts of intact temperate rainforest, once abundant runs of Pacific salmon, and the northern or southern limits of many species. Globally unique hexactinellid sponge reefs lie in deep troughs of the continental shelf. Several endemic species of plants and animals and an endemic subspecies of black bear occur on Haida Gwaii/QCI; and an unusual white form of the black bear—the Kermode or Spirit bear—lives on Princess Royal Island.

The region includes the traditional territories of 26 First Nations (aboriginal peoples) in four linguistic groups: Haida (2), Coast Tsimshian (5), Heiltsuk-Wuikala (3), Coast Salish (1), and Kwakwala (14). The total aboriginal population is about 26,000: 11,000 on reserve, 15,000 off reserve (some in the region, others in metropolitan centres in the south). Other communities in the region have a total population of about 90,000—only three with populations greater than 10,000 (Prince Rupert and Kitimat on the mainland and Campbell River on Vancouver Island).

The economy is dominated by logging, followed by the public sector, tourism, fishing, and (to much lesser degrees) aquaculture and mining. Fishing plays a major role in the subsistence economy. Economic development is hampered by limited infrastructure. Fish and marine invertebrate stocks have been reduced. Uncertainty surrounds the development of minerals, and exploitation of offshore oil and gas is barred by a moratorium. Tourism is below potential. The forest industry faces markets shrunk by the downturn in the North American economy and high import tariffs by the United States (its main market). As a result, the regional economy is in a severe recession and unemployment is high.



Page 26

First Nations in the region assert their aboriginal rights and title, which are acknowledged by the Canadian constitution and federal and provincial governments but (except for the Nisga’a) have yet to be translated into treaty settlements. Pending resolution of outstanding land issues, several Nations are preparing their own land use plans, either independently or as part of interim measures agreements with the provincial government.

1.2.2 Haida Gwaii, Central Coast, and North Coast British Columbia

Over 95% or the BC coast is designated Crown land and is managed by the Government of British Columbia. Decisions regarding land use are made via Land and Resource Management Plans (LRMPs), three of which are currently underway in the CIT region—Central Coast, North Coast, and Haida Gwaii/Queen Charlotte Islands. These three LRMPs have adopted a sectoral model with a collaborative approach to reaching decisions with a small number of people at the planning table representing larger constituencies. Representatives from the environmental community, the forest industry, tourism, recreation, labour, small business forestry, local governments, the federal government and First Nations are included.

The boundaries of the LRMP plan areas (Map 1) are not ecologically based, but instead largely correspond (by default) to other administrative and/or land use planning boundaries. For instance, in the case of the Central Coast LRMP (see below), the boundary to the east incorporates a large portion of Tweedsmuir Provincial Park and corresponds with the completed Cariboo-Chilcotin land use plan and Lower Mainland IAMC planning boundaries. The western boundary extends offshore into Queen Charlotte Sound. This boundary is subject to change. The southwestern boundary corresponds to the Vancouver Island land use plan boundary. The northern boundary extends beyond the Mid-Coast Forest District to encompass Princess Royal Island and the adjacent mainland to address the Spirit Bear park proposal and related timber issues

1.2.2.1 Central Coast

The Central Coast planning area encompasses 4.8 million ha (11.8 million acres) of land, freshwater and marine area. Extending west of the Coast Mountains, it spans the mainland coast of British Columbia from Bute Inlet in the south to Douglas Channel in the north.

The region is home to over 4,400 people, mainly First Nations. Natural resource industries, including fisheries and forestry, play a primary role in the local economies and well being of communities such as Bella Bella, Shearwater, Ocean Falls, Klemtu, Bella Coola, and Oweekeno.

The Central Coast is ecologically diverse, characterized by rugged mountains, deep ocean fjords, numerous islands and alluvial valleys that reach into the interior ecosystems of the province. Coastal temperate rainforests dominate lower elevation landscapes. These forests as well as the wetlands, bogs, estuaries, and rivers found throughout the region are biologically dynamic and rich in biodiversity.

About half the area is forested, while approximately 12% contains commercial forests available for timber harvesting. Currently, 10.74% of the Central Coast Plan Area is protected (excluding marine waters). This percentage includes provincial parks, recreation areas, and ecological reserves. Large protected areas in this region include Hakai and Fiordland Recreation Areas1 and



Page 27

a major portion of Tweedsmuir Provincial Park. Smaller protected areas include such areas as Codville Lagoon Provincial Park, Broughton Archipelago Marine Park, and the Duke of Edinburgh Ecological Reserve. An Interim Land Use Plan for the Central Coast has created 22 new protection areas, classified as Candidate Protected Areas, amounting to some 1 million ha or 20% of the region. Another suite of areas remains under negotiation and are referred to as Option Areas. Option Areas are areas where the determination of future use (Operating Area or Protection Area or some other area) is postponed pending the development of the CIT’s Ecosystem-based Management Framework and completion of the CCLCRMP after December 2003. The purpose of Option Area Status is to maintain options while the EBM is finalized over the next 12 to 24 months (Map 1).

1.2.2.2 North Coast

Covering 1.7 million ha (4.2 million acres), the North Coast planning area lies to the north of the Central Coast and stretches to the town of Stewart.

More than 20,000 people live within the North Coast, most in the city of Prince Rupert, and the remainder in the communities of Port Edward, Metlakatla, Lax Kw’alaams, Kitkatla, and Hartley Bay. These communities all lie next to the sea and draw their wealth from marine resources, forestry, and tourism.

The North Coast includes similar ecosystems to those found in the Central Coast, offering a diversity of habitat types and ecological complexes. Approximately 38% of the area is forested, and about 6% is available for timber harvesting.

1.2.2.3 Haida Gwaii/Queen Charlotte Islands

Haida Gwaii/Queen Charlotte Islands lies about 90 km west of the north coast of mainland B.C. The archipelago contains 150 islands and hundreds of islets. The total land area is just over 1 million ha (2.5 million acres).

About 6,000 people live in the Queen Charlotte Islands; a third are members of the Haida Nation. Forestry, commercial fishing, and tourism contribute to the area’s economy.

Almost 25% of the area is protected in the Gwaii Haanas National Park Reserve (147,500 ha encompassing the southern portion of Moresby Island) and Naikoon Provincial Park (73,800 ha, occupying the northeast corner of Graham Island).

1.2.3 Biological Stratification

• Ecoregional definitions are often used to delineate boundaries for conservation design and planning (Groves et al. 2000). The Ecoregion Classification system in common use in Canada stratifies British Columbia’s terrestrial and marine ecosystem complexity into discrete geographical units at five levels: Ecodomain - an area of broad climatic uniformity, defined at the global level;

• Ecodivision - an area of broad climatic and physiographic uniformity, defined at the continental level;



Page 28

• Ecoprovince - an area with consistent climatic processes and relief defined at the sub-continental level;

• Ecoregion - an area with major physiographic and minor macroclimatic variation defined at the regional level;

• Ecosection - an area with minor physiographic and macroclimatic variation, defined at the sub-regional level.

The two highest levels, Ecodomains and Ecodivisions, are very broad and place British Columbia globally. The three lowest levels, Ecoprovinces, Ecoregions, and Ecosections are progressively more detailed and narrow in scope and relate segments of the province to one another. They describe areas of similar climate, physiography, oceanography, hydrology, vegetation and wildlife potential. Within each terrestrial ecoregion, climatic zones occur where specific soils, plant and animal communities and aquatic systems develop because of the interaction of climate with the land surface and surficial materials. These zones are best defined within the BC biogeoclimatic ecosystem classification system.

The CIT study area falls within the Coast and Mountains Ecoprovince of BC. This Ecoprovince extends from coastal Alaska to coastal Oregon. In British Columbia it includes the windward side of the Coast Mountains and Vancouver Island, all of the Queen Charlotte Islands, and the Continental Shelf including Dixon Entrance, Hecate Strait, Queen Charlotte Strait, and the Vancouver Island Shelf. The Coast and Mountains Ecoprovince consists of the large coastal mountains, a broad coastal trough and the associated lowlands, islands and continental shelf, as well as the insular mountains on Vancouver Island and the Queen Charlotte Islands archipelago.

1.2.4 The ESA Study Area Boundaries

One advantage of an ecoregional approach to planning is that it can place any landscape feature in a local, regional, or global context. A second important advantage is that species, plant communities, and other conservation targets can be considered together, within an environmental framework that shaped their evolution and continues to shape their interactions. Of course, some species, especially wide-ranging animals such as grizzly bears and salmon, are ideally considered at much broader spatial scales. A third advantage of an ecoregional approach is that it enables an assessment of abundance and distribution of biological elements within their entire ecological extent or range thereby enabling valid assessment of diversity to be conducted. Moreover, using an ecoprovince or ecoregional boundary for the ESA sweeps a vast amount of geography into the assessment, over which the land use decisions that this report is meant to inform have no jurisdiction. Indeed, creating an ecological assessment that is of compatible scale and geography to the Central Coast, North Coast, and QCI/Haida Gwaii LRMP boundaries is essential if the ESA is to be used by these decision-makers. To balance these competing (but not mutually exclusive) criteria, the ESA has taken a modified approach to delineating an ecologically based study area boundary. Within the Coast and Mountains Ecoprovince, the finer-level stratification of ecosections is used to set study area boundaries. More specifically, within



Page 29

the Coast and Mountains Ecoprovince, those ecosections that intersect with the Central Coast, North Coast, and QCI/Haida Gwaii LRMP boundaries constitute the study area1 (Map2).

1.2.5 Study Area Ecoregions and Ecosections

The study area overlaps with a total of 5 terrestrial based ecoregions encompassing 10 ecosections. A sixth ecoregion is dominated by the open water marine environment and can be further divided into 4 marine ecosections. These stratifications are described as follows:

• The Coastal Gap Ecoregion contains somewhat rounded mountains with lower relief than mountain ranges to either the north or south. Valley sides are rugged and steep. Because of their lower relief, the mountains allow considerable moisture to enter the interior of the province. The Ecoregion contains two Ecosections:

– The Hecate Lowland Ecosection is an area of low relief, consisting of islands, channels, rocks, and lowlands adjacent to Hecate Strait and Queen Charlotte Sound.

– The Kitimat Ranges Ecosection is an area of subdued, yet steep-sided mountains, east of the Hecate Lowlands Ecosection.

• The Hecate Continental Shelf Ecoregion is the shallow oceanic area offshore of the Hecate Lowlands, south of the Alaska Panhandle and north of Vancouver Island. Much of this shelf lies leeward of the Queen Charlotte Islands. It contains four Ecosections.

– The Dixon Entrance Ecosection is located between northern Graham Island and Prince of Wales and Dall Islands in southeastern Alaska. This Ecosection has a strong freshwater discharge influence from the Skeena, Nass, and other rivers.

– The Hecate Strait Ecosection is a broad semi-enclosed estuarine waterway located between the mainland coast, the Queen Charlotte Islands, and northern Vancouver Island. This is a very shallow strait dominated by coarse bottom sediments. It has semi-protected waters with strong tidal currents that promote "mixing.”

– The Queen Charlotte Sound Ecosection is a deeply dissected shelf area with several large intervening banks. This Ecosection is exposed to oceanic waves allowing for oceanic water intrusions.

– The Queen Charlotte Strait Ecosection is shallow marine area that is interspersed with many islands and reefs, located between northern Vancouver Island and the Hecate Lowland. There are strong currents mixing the oceanic and freshwaters.

• The Northern Coastal Mountains Ecoregion is a rugged, largely ice-capped mountain range that rises abruptly from the coast. It contains three Ecosections in British Columbia.

• Southern Alexandria Archipelago

1 Although one of these ecosections extends into southeast Alaska, constraints of time and data availability restrict the CIT analysis to British Columbia. However, CIT data and analysis are currently being shared with a transboundary assessment—the Coastal Forests and Mountains ecoregional assessment, being performed by Round River Conservation Studies, The Nature Conservancy of Alaska, and the Nature Conservancy of Canada. This transboundary effort has an expected completion date of December 2003 and will seamlessly integrate the CIT ESA results.



Page 30

– The Alaska Panhandle Mountains Ecosection is an area of wet rugged mountains. It is the southern, windward mountain segment of the three major units in this Ecoregion. Most of this Ecosection occurs in Alaska. In British Columbia this unit occurs in the areas of Portland and Observation inlets the lower Nass River.

– The Boundary Ranges Ecosection is a large block of rugged, ice-capped, granitic mountains that are dissected by several major river valleys. It is the eastern or interior-most section of the three major units in this Ecoregion. Most of this unit occurs in British Columbia.

• The Pacific Ranges Ecoregion is the southernmost mountain range of the Coast Mountains in British Columbia. It includes the coastal islands, channels, and fjords east of Queen Charlotte Sound; otherwise it lies east of the Georgia Depression Ecoprovince. The mountains are characteristically high and rugged. It contains four Ecosections, only two of which overlap with the Study Area LRMP boundaries,

– The Northern Pacific Ranges Ecosection is an area of steep, rugged, often ice-capped, mountains located in the northern portion of this Ecoregion.

– The Outer Fiordland Ecosection is an area of rugged, low relief, consisting of inlets, sounds, islands, and peninsulas, east of Johnstone Strait and Seymour Narrows.

• The Queen Charlotte Lowland Ecoregion is represented by only one Ecosection.

– The Queen Charlotte Lowland Ecosection is an area of low relief, poor drainage and extensive muskegs and wetlands in the northeastern part of the Queen Charlotte Islands.

• The Queen Charlotte Ranges Ecoregion includes the fjords and mountains of the Queen Charlotte Mountains. Precipitation is somewhat reduced here. This Ecoregion is represented by two Ecosections.

– The Skidegate Plateau Ecosection is a plateau in the lee of the Queen Charlotte Mountains. Precipitation is lower here.

– The Windward Queen Charlotte Mountains Ecosection is the very wet and rugged western side of the Queen Charlotte archipelago.

1.2.6 ESA Ecological Drainage Units

In addition to using terrestrial based classifications for stratifying the study area, the ESA planning team delineated broad Ecological Drainage Units (EDUs) as part of their freshwater ecosystem classification framework. EDUs are groups of watersheds that share common zoogeographic history, physiographic, and climatic characteristics. EDUs contain sets of freshwater ecosystem types with similar patterns of drainage density, gradient, hydrologic characteristics, and connectivity. Identifying and describing EDUs allows us to stratify the study area into smaller units enabling us to better evaluate patterns of freshwater ecosystem diversity. Additionally, EDUs provide a means to stratify the study area to set conservation goals. A total of five EDUs were defined within the CIT region based on Hocutt and Wiley (1986), Haas’ bioregions (1998), and MacPhail and Carveth (1994) (Map 3):



Page 31

• Nass

• Skeena

• North Coast

• Central Coast, and

• Haida Gwaii/QCI.

1.2.7 Study Area Ecological Description

The moist, cool climate, mountain and lowland physiographies, and non-random natural disturbances create patterns of ecosystems on the landscape. In the Lowlands, patches of productive forest often occur within a matrix of non-forested bog lands and stunted bog woodland. In the watersheds of the Insular and Coast Mountain ranges, the forest mosaic typically consists of large, continuous tracts of all-aged, structurally diverse old-growth coastal western hemlock forest, separated by cliffs, gullies, wetlands, and shrub-covered avalanche and landslide tracks.

Within a climatically and physiographically homogeneous sub-region, slope position and substrate influence soil moisture and nutrient regime, and consequently, productivity and species composition. The biogeoclimatic ecosystem classification system identifies and interprets forested ecosystems in the CIT region (Banner et al. 1993). Where water accumulates, the wet, cool climate has led to areas of extensive unproductive bog ecosystems. Where water drains freely, the history of infrequent, stand-replacing disturbances and wet, mild climate has led to productive forests characterised by large, old trees, and tremendous accumulations of biomass, included downed wood and snags (Pojar and MacKinnon 1994). Large trees, snags, and downed wood play significant ecological roles in the CIT region, providing habitat and influencing hydroriparian processes.

On the wet, cool coast, freshwater ecosystems are particularly prevalent. In the Lowlands, wetlands (bogs, ponds and small lakes) cover 51–75% of the landscape (Banner et al. 1986, 1988). Small low-gradient streams are very common, draining the extensive slope/blanket bogs. There are many small, but few large, estuaries and floodplains, because watersheds are small and primarily rain-fed (McKenzie et al. 2000). Exposed marine shores, some supporting unique ecosystems, are common. The Insular and Coast Mountains contain a variety of freshwater ecosystems, including small, steep headwater streams and gullies, running down into fans and wide floodplains. Moderately sized linear lakes head some valleys and a variety of small wetlands dot floodplains. Large estuaries, fed by rivers, rain, glaciers, and permanent snow, are common (MacKenzie et al. 2000).

Marine Zonation - There is a strong estuarine gradient across study area, from the freshwater discharges into fjords, across the protected continental shelf to the outer continental shelf. Fjord zones are very common; nearly all large rivers empty into fjords, rather than directly onto the continental shelf. A nearshore zone surrounds all the islets, islands and mainland, with a strong intertidal zone as the dominant interface between land and sea. Extreme wind and wave exposure occurs on the west coast of the Queen Charlotte Islands, whereas more protected coasts



Page 32

occur in the Dixon Entrance, Hecate Strait, Queen Charlotte Strait and inshore areas. Most of the continental slope is dominated by mesopelagic zonation with a surface, epipelagic layer.

Fauna - Mountain goats are widespread but restricted to rugged areas in the Coast Mountains. Black bears occur throughout the region, wolves are absent from the Queen Charlotte Islands; cougars are absent from the Boundary Ranges and Queen Charlotte Islands, while grizzly bears occur only on the mainland except in the south where they have been extirpated. The sea otter was once one of the most abundant shellfish predators, and the river otter is still numerous and very widespread. Northern sea lions and harbour seals occur along the coastal areas and the killer whale is a common inhabitant. Characteristic small terrestrial mammals include the Keen’s myotis, and mink. There are many distinct island races of Townsend’s vole and white-footed mouse.

The Coast and Mountains Ecoprovince holds the second highest number of birds in British Columbia, supporting 79% of all species known to occur in the province and 57% of those species known to breed. Waterbirds make extensive use of the coastal wetlands as well as nearshore and offshore habitats, including islands, islets, and cliffs. The colonial breeding seabirds are of note, and many of those species breed nowhere else in Canada. Offshore habitats provide feeding sites for pelagic birds like the Black-footed Albatross, Sooty Shearwater, jaegers, Northern Fulmar, gulls, and some shorebirds. Breeding Red-throated Loons and Spotted Owls are mostly restricted to this Ecoprovince. Some resident species, including the Bald Eagle, Peregrine Falcon, and Black Oystercatcher, contain significant portions of their world populations here. In winter, the estuaries and shores support most of the world’s population of Trumpeter Swans and Barrow’s Goldeneyes. The coast is also an important corridor for millions of migrating birds, especially shorebirds and waterfowl. The Towsend’s Warbler is a high-density breeder on the Queen Charlotte Islands. The Western Flycatcher is also a high-density breeder on the Queen Charlotte Islands.

The centre of abundance of the northwestern garter snake occurs here. The rough-skinned newt, northwestern salamander, western red-backed salamander, ensatina, clouded salamander, and red-legged frog are amphibians whose range is mostly restricted to the Coast and Mountains Ecoprovince.

This Ecoprovince supports a wide variety of fish, from purely oceanic species such as rockfish, sole, Pacific herring, Pacific halibut and spiny dogfish, to fish that spawn in freshwater, but live as adults in marine waters, such as the Pacific salmon, steelhead, coastal cutthroat trout and eulachon, through to the species that only live in fresh water, such as Coast Range and torrent sculpins. In addition to fish the marine environment supports a wide variety of clams, barnacles, shrimp, crabs, starfish and jellyfish

1.2.8 The Coast Information Team

The Coast Information Team (CIT) has been established to provide independent information on the region using the best available scientific, technical, traditional, and local knowledge. It was set up by the Provincial Government of British Columbia, First Nations of the region, environmental groups, and forest products companies. It is led by a management committee consisting of representatives of these bodies; and is funded by the provincial government, the environmental groups and forest products companies, and the Government of Canada. The technical team



Page 33

comprises nine project teams consisting of scientists, practitioners, and traditional and local experts, supported by a secretariat. The secretariat includes an executive director, a project manager, and other part-time staff.

The CIT’s information and analyses are intended to assist First Nations and the three ongoing subregional planning processes to make decisions that will achieve ecosystem-based management (EBM), defined as “an adaptive approach to managing human activities that seeks to ensure the coexistence of healthy, fully functioning ecosystems and human communities”. The provincial government runs two of the subregional processes; the province and the Haida run the other jointly. They involve governments (First Nations, provincial, local) and a wide range of sectors (environment, fishing, forestry, labour, mining, recreation, small business, tourism). These communities and sectors are the CIT’s “stakeholders,” consultations with whom are organized by the planning processes concerned.

1.2.8.1 Components of the CIT

The CIT began work in January 2002 and was completed in May 2004. Its program consists of three types of information:

1. Ecosystem-Based Management (EBM) guides

An EBM Framework defines EBM, sets out principles to guide its implementation, and sets out goals, objectives, and key elements of EB planning and the transition to EBM.

2. Spatial analyses

An Ecosystem Spatial Analysis (ESA) identifies key areas for biodiversity conservation, focusing on representation of land, freshwater, and marine ecosystem types; protection of special elements such as rare or at-risk species and places; and maintenance of the habitats of focal species such as grizzly bear, marbled murrelet, and salmon.

A Cultural Spatial Analysis identifies key areas for sustaining the cultural values of First Nations and other communities, notably sustenance, heritage, spiritual, and recreational values.

An Economic Gain Spatial Analysis identifies key areas for economic development, focusing on fisheries and aquaculture, minerals, nontimber forest products, timber, and tourism.

3. Integrated analyses

A Wellbeing Assessment measures current environmental and human conditions in each of the eight subregions that make up the CIT analysis region, to provide a context for decision-making, a test of options and scenarios, and a baseline for monitoring implementation of the plans and progress toward EBM and sustainability. It also places the region in a global context and identifies features of global significance.

Options and Scenarios identifies combinations of key areas for ecosystem protection, cultural values, and economic gain that have the highest probability of maintaining ecosystem integrity and improving human wellbeing.



Page 34

An Institutional Analysis will examine institutional needs and constraints and additional actions required to achieve the goals of EBM.

1.2.9 The CIT Ecosystem Spatial Analyses

The purpose of the ESA is to identify priority areas for biodiversity conservation. The analysis is designed to serve the well-accepted goals of conservation summarized earlier. In pursuit of these goals, the ESA will integrate three basic approaches to conservation planning:

• Representation of a broad spectrum of environmental variation (e.g., vegetation, terrestrial abiotic, and freshwater and marine habitat classes).

• Protection of special elements: concentrations of ecological communities; rare or at-risk ecological communities; rare physical habitats; concentrations of species; locations of at-risk species; locations of highly valued species or their habitats; locations of major genetic variants.

• Conservation of habitats of focal species, whose needs help planners address issues of habitat area, configuration, and quality. These are species that (a) need large areas or several well connected areas, or (b) are sensitive to human disturbance, and (c) for which habitat suitability models are available or can be constructed.

For the CIT ESA, our challenge was to conduct an analysis of special elements, ecosystem representation, and focal species, and create a spatially explicit assessment of where the region’s biodiversity values are located and what condition they are in. This information can then be used to create a conservation solution or “portfolio” of landscapes and seascapes, which when taken together and managed appropriately, will ensure the long-term survival of the region’s biodiversity. To perform this assessment, the three-track approach was applied to freshwater, terrestrial, and marine environments using the following process:

1. Select conservation targets (e.g., special elements, focal species and ecological systems) that will be used to characterize the biodiversity values within the study area. These targets are essentially surrogates for overall biodiversity, which cannot be measured in its entirety.

2. Collect data for special element occurrences, develop habitat suitability models for focal species, and create ecosystem classifications that can be used to map the distribution of targets within the study area.

3. Using available data and simple models, assess the potential viability of targets and map human impacts in the region

4. Set conservation goals to serve as benchmarks for identifying conservation priorities and as initial hypotheses about the level of effort and land allocation required to conserve biodiversity.

5. Integrate information for special elements, ecosystem representation, and focal species in each of freshwater, terrestrial, and marine environments to create a spatially explicit assessment of conservation values for the study area.



Page 35

6. From that assessment, use goals and viability measures to develop options for creating a portfolio of conservation areas that will effectively conserve the region’s biodiversity in the long term.

This type of analysis employs and integrates many thousands of pieces of detailed information. It requires location-specific information for conservation targets as well as the past, current, and potential future status of lands where they occur. Our team used the best available information for this assessment but recognizes that new and more comprehensive data will continually become available. Therefore, the ESA should be regarded as an initial step in an iterative assessment process.



Page 36

2.0 CONSERVATION TARGETS

2.1 The Three Track Approach to Setting Conservation Targets and Goals

Over the last few decades, many scientific approaches have been applied to the task of identifying conservation targets and setting conservation goals. Most of these approaches fall into the three general categories or tracks noted earlier as forming the basis of the comprehensive approach applied in this ESA: 1) protection of special elements; 2) representation of environmental variation, i.e., habitats; and 3) conservation of focal species.

Each of these three approaches—or even different ways of conducting a given approach—arrives at a unique set of conservation priorities, which are often difficult to reconcile with the priorities established by other methods. Someone interested in rare plants, for instance, will arrive at different conservation priorities from someone interested in songbirds; both will differ in their conclusions from someone interested in representing examples of all plant communities in reserves or maintaining a viable population of grizzly bears. The data sought to fulfill these goals also will vary greatly in type, spatial scale and resolution, and completeness. Few previous regional conservation plans have combined all three tracks and their associated data, yet such a combination is necessary to make fully informed decisions about land and water allocation and management (see Noss et al. 2002). Moreover, applying a diversity of approaches in conservation planning spreads the risk of failure of any single approach and potentially achieves a more comprehensive set of goals (Lindenmayer et al. 2002; Noss et al. 2002).

2.1.1 Special Elements

The special elements approach typically results in the mapping of hotspots and other biologically or ecologically important areas that are recommended for protection above other areas. Hotspots usually are based on concentrations of species (usually rare or endemic taxa) and can be recognized on a variety of spatial scales, from locally to globally (e.g., see Myers et al. 2000). Identified hotspots of species richness or endemism, and any other priorities based on special elements, are only as reliable as the underlying data. In most cases, including the majority of British Columbia and the rest of Canada, biological surveys are spotty at best. Areas that show up as “cold spots” could either be areas where species richness or endemism is truly low or they could simply be areas that were never surveyed.

The Nature Conservancy (U.S.) and the Nature Conservancy of Canada traditionally emphasized a special elements approach, often referred to as the “fine filter” (Noss 1987), although they have today expanded to an ecoregional planning strategy as summarized earlier (Groves et al. 2002, Groves 2003). In the fine filter, individual occurrences of imperiled species (which may or may not correspond to populations), communities, and other features are located, mapped, and targeted for protection. The fine-filter approach works well for plants and small-bodied animals, especially in regions where biodiversity databases (e.g., Conservation Data Centres) are reasonably complete. It is not as well suited for large-bodied or wide-ranging animals, such as grizzly bears, salmon or northern goshawks, whose needs cannot be captured by occurrence data. In all cases, the fine filter is dependent on reasonably comprehensive, or at least well-distributed, biological surveys to be most useful. Although surveys are not comprehensive for most of



Page 37

Canada, to neglect areas known to be rich in special element occurrences or other ecological values simply because survey data across the region in question are incomplete would be foolhardy. A precautionary approach would protect known hotspots. Hence, the fine filter remains valuable (indeed necessary, if not sufficient) even in relatively poorly surveyed regions.

2.1.2 Representation

In contrast to the fine-filter or special elements approach, the “coarse filter” is intended to protect high-quality examples of all ecosystems in a region. If applied to small, localized occurrences of imperiled natural community types, as it often has been in practice, the coarse filter is really no different from the fine filter or special elements approach. If applied on a landscape scale, however, with the notion of representing all ecosystems in a region across their natural range of variation along environmental gradients, the coarse filter is complementary to special elements conservation (Noss 1987, Hunter et al. 1988, Hunter 1991).

The coarse filter is an example of the representation strategy, the history of which extends back to the late 19th century in Australia and the early 20th century in North America (Noss and Cooperrider 1994; Scott 1999). One of the strongest arguments for the representation strategy is that it is likely to capture species, genes, communities, and other elements of biodiversity that are poorly known or surveyed. Bacteria, fungi, bryophytes, and many invertebrate groups, for instance, would rarely be considered in the special elements track, simply because data on their distributions are not available. In a sense, the coarse filter serves as a buffer for our ignorance about biogeography (Hunter 1991).

Given that species distributions are determined largely by environmental factors, such as climate and substrate, and that vegetation and other species assemblages respond to gradients of these factors across the landscape, protecting examples of all types of vegetation or physical environmental classes ought to capture the vast majority of species without having to consider those taxa individually (Noss and Cooperrider 1994). It has been estimated that 85-90% of all species can be protected by the coarse filter (Noss 1987). Testing this optimistic assumption empirically is difficult, as doing so would require a reasonably complete inventory of all taxa, including cryptic organisms such as bacteria and small invertebrates, sampled over a broad area. In Victoria, Australia, vegetation classes represented birds, mammals, and trees fairly well, but performed poorly for reptiles and invertebrates (MacNally et al. 2002). In regions with relatively low endemism, such as most of Canada, the coarse filter is predicted to perform better than in regions with high endemism, where species populations are highly localized (Noss and Cooperrider 1994).

Representation assessments typically rely on vegetation (often mapped by remote sensing, as in the U.S. Gap Analysis Program; Scott et al. 1993), surrogate taxa (e.g., vertebrate species richness, also used in Gap Analysis), abiotic environmental classes (e.g., landforms, habitat classes defined by soils or geology), or some combination of biological and physical factors (e.g., ecological land units) as proposed coarse filters. Increasing evidence suggests that a combination of biological and abiotic data, as in ecological land units, provides a more secure basis for representation than either class alone (Kirkpatrick and Brown 1994; Kintsch and Urban 2002; Noss et al. 2002a; Groves 2003; Lombard et al. 2003).



Page 38

2.1.3 Focal Species

For the ESA, the focal species approach can be distinguished from the species component of the special elements track, in that habitat suitability are modelled and extrapolated beyond currently known occurrences. Modelling approaches for focal species may be either static or dynamic; for greatest utility, both types of models should be spatially explicit and GIS-based. Static habitat suitability models, such as those employed in this study, can be either conceptual models, based on existing literature and expert opinion about species-habitat relationships, or empirical models, e.g., resource-selection functions built by associating occurrence data with potential predictor variables through multiple logistic regression or other statistical techniques (Manly et al. 1993; Boyce and McDonald 1999; Carroll et al. 2001). Data needs for empirical models include an adequate sample of point occurrences for the species in question and GIS databases on a variety of potential predictor variables, including vegetation, topography, climate, prey availability (or a surrogate thereof), metrics derived from satellite imagery, and human-impact variables such as population density and roads (Carroll et al. 2001; Noss et al. 2002; Carroll et al. 2003).

With further work that currently lies beyond the scope of the CIT ESA, focal species analysis can complement the special elements and representation tracks by addressing questions concerning the size and configuration of reserves and other habitats necessary to maintain viable populations. It is highly recommended that future work through an EBM framework concentrate on using these habitat models to create connectivity and population viability analysis.

A further advancement would be the use of dynamic population models that may include several kinds of spatially explicit population viability analyses (PVAs) (Beissinger and McCullough 2002). These models are typically more time-consuming and difficult to construct than static models, and require more information on the demographic characteristics of each focal species. Hence, our focal species models do not explicitly consider population viability. We recommend that dynamic population models for selected species be developed in the future.

2.2 Terrestrial Targets


2.2.1.1 Methods

Special element (“fine filter”) targets were selected based on global, national, and provincial conservation status within the larger ecological boundaries of the Coast and Mountains Ecoprovince (COM).2 Also targeted were “Species of Special Concern”—species or subspecies which globally are apparently secure and/or abundant (ranked G3–G5 by BC Conservation Data Centre and NatureServe), but within the COM exhibit the following characteristics: exhibit significant, long-term declines in habitat and/or numbers, are subject to a high degree of threat, or may have unique habitat or behavioural requirements that expose them to great risk; are restricted to the COM (or a small geographic area within the ecoprovince), depending entirely on the ecoprovince for survival, and therefore may be more vulnerable than species with a broader distribution; have populations that are geographically isolated from other populations; are more widely distributed in other ecoprovinces but have populations in the COM at the edge of their

2 For an overview and description of the Coast and Mountains Ecoprovince refer to BC MSRM webpage: http://srmwww.gov.bc.ca/ecology/ecoregions/humidtemp.html#coast



Page 39

geographical range; are usually abundant and may or may not be declining, but some aspect of life history makes them especially vulnerable – e.g., migratory concentration or rare/endemic habitat; have spatial, compositional, and functional requirements that may encompass those of other species in the region and may help address the functionality of ecological systems; are unique, irreplaceable examples for the species that use them, or are critical to the conservation of a certain species or suite of species; are critical migratory stopover sites that contain significant numbers of migratory individuals of many species (TNC 2000; Comer 2001; Groves et al. 2002). Table 2.1. summarizes the details of target selection criteria.

2.2.1.1.1 Data collection

A database was created with information on species and communities obtained from BC Conservation Data Centre (BC CDC), BC Ministry of Forests, Committee On the Status of Endangered Wildlife In Canada (COSEWIC), Partners In Flight, and NatureServe databases, a review of BC coastal land use planning documents, pertinent research (e.g., Calder and Taylor 1968; Campbell et al. 1990; Cannings and Ptolemy 1998; Lomer and Douglas 1998; Cannings et al. 1999; Douglas et al. 2002), and interviews with species and communities experts. Staff from the BC Conservation Data Centre, the BC Ministry of Water, Land and Air Protection, the BC Ministry of Sustainable Resource Management, BC Ministry of Forests, and the Royal BC Museum then reviewed this database. As well, target lists were reviewed by planning team members.



Page 40

Table 2.1 Special elements target selection criteria

Criteria Rank Description

Global conservation status

G1–G3; T1–T3

Subnational (provincial) conservation status

S1–S3

1 = Critically Imperilled either because of known threats or declining trends, or because extremely restricted breeding or non-breeding range make the element vulnerable to unpredictable events, a candidate for “endangered” status; 2 = Imperilled, a candidate for “threatened” status; 3 = Vulnerable – usually more abundant or widespread than 1 or 2, but sensitive to threats, perhaps declining (BC CDC, NatureServe)

National conservation status (Committee On the Status of Endangered Wildlife In Canada)

E

T

SC

Endangered (E) – A species facing imminent extirpation or extinction.

Threatened (T) – A species likely to become endangered if limiting factors are not reversed.

Special Concern (SC) – A species that is particularly sensitive to human activities or natural events but is not an endangered or threatened species (COSEWIC 2003).

Provincial listing

(BC Conservation Data Centre)

Red

Blue

Red – includes any indigenous species or subspecies that have, or are candidates for Extirpated, Endangered, or Threatened status in British Columbia. Extirpated taxa no longer exist in the wild in British Columbia, but do occur elsewhere. Endangered taxa are facing imminent extirpation or extinction. Threatened taxa are likely to become endangered if limiting factors are not reversed.

Blue – includes any indigenous species or subspecies considered to be of Special Concern (formerly Vulnerable) in British Columbia. Taxa of Special Concern have characteristics that make them particularly sensitive or vulnerable to human activities or natural events. Blue-listed taxa are at risk, but are not Extirpated, Endangered, or Threatened.

Partners In Flight Score (for Bird Conservation Region 5 – Northern Pacific Rainforest)

Sum of Vulnerability Factors.

Scores for each factor range from 1 (low vulnerability) to 5 (high vulnerability).

Relative Abundance – reflects the abundance of breeding individuals of a species, within its range, relative to other species; Breeding Distribution – reflects the global distribution of breeding individuals of a species during the breeding season; Non-breeding Distribution – reflects the global distribution of a species during the non-breeding season; Threats to Breeding – reflects the effects of current and future extrinsic conditions on the ability of a species to maintain healthy populations through successful reproduction. Threats to Non-breeding – reflects the effects of current and future extrinsic conditions on the ability of a species to maintain healthy populations through successful survival over the non-breeding season; Population Trend – reflected by the direction and magnitude of changes in population size over the past 30 years; Area Importance – reflects the relative importance of an area to a species and its conservation, based on the abundance of the species in that area relative to other areas.

Special Concern Declining

Endemic

Disjunct

Peripheral

Vulnerable

Umbrella species

Species aggregations

Declining - exhibit significant, long-term declines in habitat and/or numbers, are subject to a high degree of threat, or may have unique habitat or behavioural requirements that expose them to great risk; Endemic - are restricted to the COM (or a small geographic area within the ecoprovince), depending entirely on the ecoprovince for survival, and therefore may be more vulnerable than species with a broader distribution; Disjunct - have populations that are geographically isolated from other populations; Peripheral - are more widely distributed in other ecoprovinces but have populations in the COM at the edge of their geographical range; Vulnerable - are usually abundant and may or may not be declining, but some aspect of life history makes them especially vulnerable – e.g., migratory concentration or rare/endemic habitat; Umbrella species - have spatial, compositional, and functional requirements that may encompass those of other species in the region and may help address the functionality of ecological systems; Species aggregations - are unique, irreplaceable examples for the species that use them, or are critical to the conservation of a certain species or suite of species; Globally significant examples of species aggregations - are critical migratory stopover sites that contain significant numbers of migratory individuals of many species.



Page 41

2.2.1.1.2 Target list

The terrestrial special elements database consists of 110 targets (75 vascular plants, 9 birds, 5 mammals, and 21 rare plant communities), which met the criteria outlined previously. Spatial data was obtained for 72 targets. Table 2.2 provides a summary of the final terrestrial fine filter targets database.

Table 2.2 Summary of terrestrial fine filter targets

Taxa

Targets with spatial

data/

Total # targets

# of targets ranked

G1–G3; T1–T3 Endemics

# of targets ranked

Red or blue (BC CDC)

# of targets ranked

endangered, threatened, or special Concern

(COSEWIC)

# of element

occurrence records

Birds

5/9 7 8 9 1 150

Mammals 3/5 3 2 5 4 5

Vascular Plants

52/75 20 9 75 205

Rare plant communities 12/21 21 21 10

TOTAL 72/110 30 40 110 5 370

2.2.1.1.3 GIS data

Spatial data was collected from all available sources. The majority came from the BC Conservation Data Centre (>90%), which collects and maintains data on rare and endangered species in the province (Map 4). The data sets were screened based on data quality. CDC element occurrence records ranked as “low quality,” extinct, extirpated, or historic (BC CDC codes D, E, EI, H, X, and X?) were removed from the final GIS data set. As well, animal occurrence data >25 years and plant occurrence data >40 years were deleted (TNC 2000; Rumsey et al. 2003). Appendix 1.1.1 details the complete list of special elements.

2.2.1.2 Results and Discussion

Spatially, most of the special elements data are located on Haida Gwaii/QCI where there are 227 element occurrence records. The North Coast contains 49 records and the Central Coast has 94 records. Most of these special element occurrences are vascular plants and rare plant communities. Calder and Taylor (1968) surveyed Haida Gwaii/QCI on several occasions where they identified 593 vascular plants. As well, the BC Conservation Data Centre conducted surveys in 1997 and 1998 to locate and document the size and condition of provincially rare native vascular plant populations where they increased the number of identified taxa to 665 (Lomer and Douglas 1998). Although the CDC and Ministry of Forests maintain a database of rare plant associations from biogeoclimatic mapping, there was very little spatial data on rare plant communities.



Page 42

The special elements analysis highlights the scarcity of occurrence data on much of the BC coast. The majority of element occurrence data is on Haida Gwaii/QCI reflecting an uneven distribution of surveying effort. There is also a concern that many of the occurrence points are near roads, possibly reflecting a surveying bias. Instead of using the fine filter data as an input to the SITES runs, the ESA has simply packaged occurrence data as part of its final database so as to allow makers-makers and stakeholders at LRMP tables to assess protected areas and management regimes for how well these elements are being represented.

2.2.1.3 Recommendations 2.2.1.3.1 Initial conservation goals

For future iterations of this plan and other planning efforts in the region, given the state of our limited knowledge on target viability and population dynamics, we recommend establishing initial conservation goals for special elements, then refining these goals as much as possible with target-by-target information (Comer 2001).

Initial conservation goals can be set for terrestrial targets based on their geographic scale, distribution and spatial pattern. Initial conservation goals can attempt to represent the “natural” or historic range of distribution for the target. For example, if 50% of the known, natural range of the target falls within the ecoprovince, the goal for the ecoprovince should reflect roughly 50% of a range wide goal. The target’s distribution, relative to the ecoprovince is used to establish numeric differentials in goal setting—i.e., higher with endemic, to lower with peripheral (Comer 2001, TNC 2000, Rumsey et al. 2003).

2.2.1.3.2 Data gaps

The fine filter analysis was lacking in invertebrate data. No sources for location data were found. Coastal temperate rainforests are known to have very high invertebrate biodiversity (Scudder 1996). There are also a number of rare and endangered lichens that were not included in the fine filter database. Although these are outlined in The Lichens of British Columbia (Goward 1999), we were not able to obtain or create digital files for these locations. Haida Gwaii/QCI is one of the few areas where comprehensive lichen studies have been carried out (Goward 1999). Ongoing conservation efforts through ecosystem-based management should strive to acquire such data.

2.2.2 Representation

2.2.2.1 Background and Rationale

Coastal British Columbia, a region characterized by moderate climates, high rainfall (192 cm or more), and proximity to both mountains and the Pacific Ocean (Pojar et al. 1987), contains a unique assemblage of terrestrial ecosystems, including glaciers and steep mountain systems, high elevation alpine tundra, coastal muskeg forests and woodlands, estuarine and riparian systems, intertidal and coastal habitats and old-growth coastal temperate rainforests (Meidinger and Pojar 1991). This section outlines coarse-filter spatial data and methods for representation of a full range of terrestrial ecosystem components that are found in coastal British Columbia.

The coastal temperate rainforest is a globally rare ecosystem (Smith and Lee 2000) and is highly vulnerable to continued industrial activities, therefore, identification and representation of a suite



Page 43

of old-growth ecological systems is central to this coarse-filter conservation planning approach. In recent times, old-growth coastal temperate rainforests of North America, particularly communities dominated by Sitka spruce, Douglas-fir, and western redcedar, have seen massive changes in distribution, composition and age structure (Schoonmaker et al. 1997; Smith and Lee 2000). The reason for these anthropogenic changes is not because coastal forests are exceptionally vulnerable to human disturbance but instead, the forests themselves, particularly stands that contain a large volume of old trees, are economically valuable and have been targeted by industrial scale logging. Thus, identification and protection of the best examples of remaining old-growth forests is critical to the success of long-term conservation efforts, not because forest communities are particularly sensitivity to disturbance, but rather in response to unparalleled resource exploitation in every place old-growth coastal temperate rainforest was previously found.

Coastal old-growth forest ecosystems are distinguished by late-successional plant communities and related structural features. Coastal old-growth characteristics and definitions have been the subject of intense scientific research and legal scrutiny and old growth has been described variably in terms of stand structures (Franklin et al. 1981), stand development processes (Oliver and Larson 1990) and a combination of perspectives including genetic, population, ecosystem and landscape levels (Spies and Franklin 1995). Old-growth definitions tend to include characteristics related to the later stages of stand development, that typically differ from earlier stages based on tree size, accumulations of large, dead, woody material, canopy layers, species composition, function, and other attributes (e.g., Franklin et al. 1986). These structural characteristics often include pronounced high timber volume areas containing dramatic examples of large and old trees. We used structural and age class data in a manner designed to identify a range of old-growth forest ecosystems. Unfortunately, many of the best examples of coastal temperate rainforest ecosystems have already been destroyed by industrial activities. Therefore, a quantitative consideration of levels of historic impacts and setting goals for inclusion of areas based on historical distribution is also an explicit component of this analysis and we describe a method to first represent intact ecosystems, followed by inclusion of impacted areas if necessary to meet representation goals.

Coarse-filter approaches described here are designed to identify regional-scale or system-scale biodiversity features (e.g., biogeoclimatic and general ecological system or floristic types), rather than specific, fine-scale, vegetation community types or species, assuming that the broader-scale biodiversity surrogates sufficiently represent the finer-scale aspects of biodiversity (Pressey and Logan 1994; Pressey 1994; Williams and Humphries 1996; Wessels et al. 1999; Fairbanks and Benn 2000; Fairbanks et al. 2000). Moreover, representing a full spectrum of abiotic types and associated vegetation, especially if done in large, contiguous ecologically intact areas, may facilitate shifts in species distributions in response to climate change (Noss 2001). However as Pressey (1994) points out, the assumed relationship between environmental classes and species distributions is unclear and seldom investigated. In addition, certain species, especially rare species confined to small patches of habitat which are not recognized as distinct coarse-filter classes, or which cross boundaries of coarse-filter classes may fall through the coarse-filter when using broad-scale classification techniques (Noss 1983; Bedward et al. 1992; Panzer and Schwartz 1998). To address these shortcomings, we suggest that coarse-filter approaches can be used in combination with finer-scale species and ecosystem distribution information. However, because such fine scale information is patchy, with limited spatial coverage, such data has limited utility



Page 44

for initial phases of regional conservation planning efforts, such as for driving optimization portfolio models (e.g., SITES), since areas selected would be based largely on the presence or absence of data rather than representation. We suggest that such fine-scale information can be used as an integral part of representation analysis, specifically for post-optimization, verification of representation. This two-step approach has the advantage of using the best available data in a scale-appropriate manner. This approach also allows for creating combinations of physical and floristic information sets without the need for laborious groupings into specific plant community types.

Thus, we developed two methods to identify and represent different coarse-filter components of terrestrial ecosystems. The first was based on biogeoclimatic zone, subzone and variant information combined with seral stage and site productivity information. This approach allowed us to determine the historic impact of various ecological system types and set goals based on historic abundance. The second approach was based on overstory species composition combined with a range of structural characteristics with the goal of representing both the structural and functional features that are characteristic of coastal temperate rainforests.

2.2.2.2 Methods

In BC, two regional ecosystem classification systems exist. The ecoregion/ecosection classification, developed by the B.C. Ministry of Environment, Lands and Parks, describes broad regional ecosystems based on the interaction of climate and physiography. The biogeoclimatic ecosystem classification (BEC) delineates ecological zones (biogeoclimatic units) by vegetation, soils, and climate. We used both systems and applied two classification methods, described in the sections below.

2.2.2.2.1 Terrestrial ecosystem analysis units: Biogeoclimatic, site productivity, and seral stage

The biogeoclimatic ecosystem classification (BEC) delineates ecological zones (biogeoclimatic units) by vegetation, soils, and climate (Pojar et al. 1987; Meidinger and Pojar 1991, Map 6a). We combined BEC information with site productivity (taken from Site Index in the BC forest cover data), to delineate terrestrial ecosystems, with the assumption that site productivity correlates with areas that have climax, or old growth, ecosystem characteristics (including canopy structure and understory vegetation and associated faunal community types). Because both BEC and site index rely largely on physiognomic characteristics, the relative impacts by ecosystem type can be predicted, without knowledge of the type of vegetation that was present before the areas were logged. Note that here we also assume that impacted areas of the same BEC and site index as intact areas will have similar old-growth characteristics and we also assume a relatively low level of natural disturbance (e.g., infrequent fires, Morrison and Swanson 1990) in BC coastal forest ecosystems. An advantage of application of this method is the capacity to identify matched ecosystems in both intact and impacted areas, because many of the best examples of coastal temperate rainforest ecosystems were targeted for logging because of the economic value and accessibility of particular areas, and in many cases, few examples of intact ecosystems remain (see Table 2.3). This method also allows us to preferentially identify and represent intact areas, and then represent comparable impacted areas for potential modified, only if necessary to meet ecosystem representation goals.



Page 45

Site Index was grouped into three classes (Low = 1–14, Medium = 15–21, and High > 22). BEC zones were taken from the BC biogeoclimatic classification coverage. We treated separate variants and subzones as separate units because in some areas, variants were not defined (e.g., CWHvh1 and CWHvh were treated as different BEC zones, see Table 2.3). Logging data was taken from a mixture of forest cover and Small Scale Predictive Ecosystem Mapping (SSPEM), combined with Sierra Club B.C. image analysis to determine intact and impacted areas. Because our logging data was derived from multiple sources, logged areas often overlapped with defined old forest (i.e., Age Class > 7) areas in the database. We considered such areas as logged and set Age Class = 1, and considered these areas to have been logged. Thus, seral stage is also implicit in this classification system and we defined 3 seral stages: old-growth forest (Age Class > 7), young intact forest (Age Class < 8 and unlogged), and logged forest. Historical abundance of each ecosystem type was calculated as the sum of area for all seral stages (Map 6b).

Table 2.3 Biogeoclimatic zones (based on zone, subzone, and variant) and logging impacts by site productivity

Percent logged

Biogeoclimatic zone -

subzone/variant Area (ha)

Forested area (ha w/known SI and ITG)

Low (Site Index =

1–14)

Med (Site Index =

15–22)

High (Site Index >

22)

AT 95,335.25 45,111.50 0.03% 0.00% 0.00%

CWHdm 65,501.00 18,830.00 71.01% 90.32% 72.67%

CWHds1 50,158.00 Unknown

CWHds2 68,184.00 51,608.50 7.28% 57.92% 27.23%

CWHmm1 8,855.50 7,679.00 61.00% 95.09% 85.93%

CWHms1 45,382.25 Unknown

CWHms2 112,958.75 109,917.50 11.77% 53.73% 27.62%

CWHvh1 116,426.25 103,273.00 7.50% 88.75% 45.82%

CWHvh2 1,519,127.00 1,439,084.75 0.78% 28.41% 13.25%

CWHvm1 200,350.75 130,040.75 9.59% 68.47% 37.55%

CWHvm2 807,201.50 734,927.50 1.35% 21.55% 9.37%

CWHvm3 476,753.50 419,905.25 0.89% 48.53% 6.16%

CWHvm 49,751.50 48,426.50 2.99% 35.43% 9.07%

CWHwh1 484,662.75 437,232.25 2.08% 39.31% 13.30%

CWHwh2 79,459.25 62,609.75 3.63% 67.06% 10.39%

CWHwm 126,119.75 94,017.75 0.77% 0.84% 4.40%

CWHws1 172,983.00 118,936.50 24.50% 58.99% 66.14%

CWHws2 495,181.75 318,803.25 2.83% 51.91% 23.22%

CWHxm2 49,345.00 28,792.25 92.08% 83.65% 90.10%

ESSFmc 513.00 509.50 0.25%

ESSFmcp 165.00 164.75 0.00%

ESSFmk 79,368.75 59,027.25 0.00% 0.00%

ESSFmkp 2,944.50 2,924.75 0.00%



Page 46

Percent logged

Biogeoclimatic zone -

subzone/variant Area (ha)

Forested area (ha w/known SI and ITG)

Low (Site Index =

1–14)

Med (Site Index =

15–22)

High (Site Index >

22)

ESSFmw 43,728.25 11,804.25 0.03% 0.00% 0.00%

ESSFwv 134,902.00 132,519.50 0.65% 3.07% 38.26%

ESSFwvp 9,014.75 8,828.50 0.00% 0.00% 0.00%

ESSFxv1 5,188.50 Unknown

ICHmc1 18,464.25 17,161.50 3.93% 30.14% 56.68%

ICHmc2 139,536.00 110,590.00 10.39% 15.41% 30.60%

IDFdw 3,497.75 Unknown

IDFww 1.75 Unknown

MHmm1 380,487.25 276,743.50 0.20% 7.26% 1.30%

MHmm2 393,276.50 234,213.75 0.35% 63.39% 10.25%

MHmmp 38,283.00 37,362.25 0.00% 0.00% 0.00%

MHwh1 68,487.75 62,439.25 0.10% 19.48% 1.77%

MHwh2 16,714.00 8,859.00 2.81% 50.39% 5.86%

MHwhp 4,835.50 4,668.00 0.00% 0.00% 0.00%

SBSdk 4.25 4.25 100.00%

SBSmc2 121.75 117.25 5.00% 2.25%

Goals for the site-selection algorithm were set at 30, 40, 50, 60, and 70% representation of total area for the ecosystem type based on historical abundance. However, for some ecosystem types, the representation goal exceeded the remaining unlogged forest. For example, 88.75% of medium productivity CWHvh1 has been impacted (see Table 2.3), so even a 30% representation goal would exceed the remaining non-impacted forest of this type. Representation targets were selected from areas that had both Site Index and species information (i.e., with known vegetation information), to eliminate bare rock and ice areas. We applied the following rules for goal setting:

• If a representation goal can be met using non-impacted areas (i.e., target ≤ remaining forest), no goal was set for logged areas; goals for remaining forest were based on historical abundance estimates.

• If impacts equal or exceed 85%, the remaining old-growth forest and young intact forest goals were set to 100%, and the goal for logged forest was set as the difference between the overall representation target area and the remaining intact forest area. This can be expresses as the following formula:

– Logged forest area target (ha) = representation goal/100 * historic distribution area (ha) – (old growth (ha) + non-impacted young forest (ha))

• Where impacts were less than 85%, a 95% target was set for non-impacted areas, with remaining representation target set using logged areas as above.



Page 47

2.2.2.2.2 Focal ecosystems: Floristic, structural, and ecosection combinations

To complement the Site Productivity and BEC method for identifying ecosystems, we combined floristic and structural data to identify focal ecological systems (Map 7). Several species of conifers inhabit the coastal rainforest of BC and we used Size Class, Inventory Type Group (ITG), and Age Class to define and delineate stands of old growth and woodland areas based on both floristic and structural characteristics. We grouped ITGs to identify ecological systems. Because the same species groups from the forest cover database may signify different ecological systems in different ecosections, we stratified goal setting by ecosection. For example, high volume western redcedar forests in the Hecate Lowlands probably represent different on-the-ground terrestrial ecosystems than high volume western redcedar forests in the Kitimat Ranges (or in any other ecosection).

This method allows us to represent a full range of ecosystem types without the need to know exactly which ecosystem or community type is present (this is a major advantage for this coarse-filter approach). We do this to help capture some of the structural, functional, and age characteristics of coastal B.C. terrestrial systems, including a range of old-growth forest ecosystems. Focal ecosystems were simply defined as unique combinations of structure, ecological system (as defined in Tables 2.4 and 2.5) and ecosection. Goals for focal ecosystems were varied from 30 to 70% in 10% increments.

Table 2.4 Ecological system and alliance species groupings based on ITG

ITG ITG_desc Alliance Ecological system

1 Fd Douglas-fir Douglas-fir Forest

2 FdCw Douglas-fir Douglas-fir Forest

3 FdH Douglas-fir Douglas-fir Forest

4 FdS Douglas-fir Douglas-fir Forest

5 FdPl Douglas-fir Douglas-fir Forest

6 FdPy Douglas-fir Douglas-fir Forest

7 FdL Douglas-fir Douglas-fir Forest

8 FdDecid Mixed Doug-Fir–Deciduous Mixed Deciduous Forest

9 Cw Cedar Cedar Forest

10 CwFd Cedar Cedar Forest

11 CwH Hemlock–Cedar Hemlock–Cedar Forest

12 H Hemlock Hemlock Forest

13 HFd Hemlock Hemlock Forest

14 HCw Hemlock Hemlock Forest

15 HB Hemlock–Silver Fir Hemlock–Silver Fir Forest

16 HS Hemlock–Spruce Hemlock–Spruce Forest

17 Hdecid Mixed Hemlock–Deciduous Mixed Deciduous Forest

18 B Hemlock–Silver Fir Hemlock–Silver Fir Forest

19 BH Hemlock–Silver Fir Hemlock–Silver Fir Forest



Page 48

ITG ITG_desc Alliance Ecological system

20 BS Hemlock–Silver Fir Hemlock–Silver Fir Forest

21 S Spruce Spruce Forest

22 SFd Spruce Spruce Forest

23 SH Hemlock–Spruce Hemlock Spruce Forest

24 SB Spruce Spruce Forest

25 SPl Spruce–Pine Spruce–Pine Forest

26 Sdecid Mixed Spruce–Deciduous Mixed Deciduous Forest

27 Pw Western White Pine

28 Pl Shore Pine Lodgepole Pine Forest

29 PlFd Shore Pine Lodgepole Pine Forest

30 PlS Spruce–Pine Spruce–Pine Forest

31 PlDecid Mixed Shore Pine–Deciduous Mixed Deciduous Forest

32 Py Ponderosa Pine

33 LFd Larch Deciduous Forest

34 L Larch Deciduous Forest

35 AcConif Mixed Cottonwood–Conifer Riparian Forest

36 AcDecid Cottonwood Riparian Forest

37 DrDecid Alder Deciduous Forest

38 Mb Maple Deciduous Forest

39 E Birch Deciduous Forest

40 AtConif Mixed Aspen–Conifer Mixed Deciduous

41 AtDecid Aspen Deciduous Forest

Table 2.5 Focal ecological system structure classes

Height class Age class Structure

> 4 >7 High Volume Old Growth

= 3,4 >7 Med Volume Old Growth

< 3 Any Intact Woodland

2.2.3 Focal Species: Marbled Murrelet (Brachyramphus marmoratus) Nesting Habitat Suitability


The Marbled Murrelet (Brachyramphus marmoratus) is a small brown alcid whose biology has given it an unusually important role in resource management in British Columbia. Unlike other auks it occurs regularly throughout all of the nearshore waters of coastal BC. Where currents or tidal rapids concentrate food, it may occur in the hundreds. It is the only auk to be found in large



Page 49

numbers (200–700) in the coastal fiords and is often the most abundant water bird in those areas. The murrelet usually feeds on small marine fish, such as sand lance (Ammodytes hexapterus), juvenile Pacific herring (Clupea harengus), northern anchovy (Engraulis mordax), and immature rockfish (Sebastes spp.), but it will also take large zooplankton, such as euphausiids (Burger and Chatwin 2002). Current population estimates for British Columbia are about 76,000 birds. The BC population is much larger than those in Washington, Oregon, and California combined but less than 5% of the Alaska population.

In spite of their apparent abundance and wide distribution, the murrelet is the only seabird listed by the Committee On the Status of Endangered Wildlife In Canada (COSEWIC). It is also Red-listed by the BC Conservation Data Centre, and is an Identified Wildlife Management Strategy (IWMS) species under the Forest Practices Code. That paradox between abundance and the conservation status stems from features of the murrelet’s breeding biology that have only recently come to light and are still not thoroughly understood. Unlike most other alcids in BC, the Marbled Murrelet does not nest colonially in underground burrows and does not nest on small offshore islands close to richer concentrations of marine resources. It nests solitarily, in trees, on the mainland, often 20 to 30 km and up to 60 km or more from the sea. However most nests are less than 5 or 10 km from salt water. The murrelet uses energetically expensive high-speed flight to make a daily commute of as much as 100 km between its nest and preferred marine foraging areas. It depends on its ability to catch calorie-dense fish such as sand lance to ignore the costs of speed and distance. It appears to be a highly effective strategy and, unlike other alcids, it does not lose weight while making the effort to feed its young.

With only a few exceptions, the nests are always in some form of old-growth forest. Radio telemetry has located several hundred nests at high elevations and on steep slopes but the murrelet also nests in the large trees of lowland forests. High elevations may help murrelets escape the attentions of ravens, jays, and flying squirrels that would take the egg or young if the opportunity arose. However, high elevation nests are a relatively recent discovery and it was the widespread loss of big-tree lowland habitat that first brought the Marbled Murrelet to the attention of the conservation community. The proportion of the population that makes use of each of those habitats, and their relative rates of reproductive success, are unknown. Because the murrelet readily commutes long distances, there is only a coarse relationship between concentrations at sea and suitable nesting habitat on land.

Murrelets prefer to nest in areas with a relatively closed canopy that has trees with more than average structural complexity. The nests are little more than shallow depressions in moss or duff on large tree limbs. They vary in elevation and aspect but typically well within the canopy and often close to the trunk of the tree. The actual nest is often partially obscured by overhanging branches. Murrelets usually approach their nests in the dark and may select sites close to highly visible (to them) landmarks such as avalanche runs, gulleys, rocky outcrops, and other forest openings.

The Marbled Murrelet was chosen as a focal species due to conservation concerns over loss of nesting habitat and increased predation risk from forestry activity. There is increasing concern for the influences of human activity on survival at sea (Steventon 2002). Marbled Murrelets face a growing host of marine-related impacts such as oil pollution, fisheries by-catch, and aquaculture (Robinson 2002). Effects of a strong increase in recreational boat traffic and the subsequent spread



Page 50

of developments such as marinas and small harbours have not been assessed. Sport fishermen snag a significant number of murrelets each year.

The secretive terrestrial life of Marbled Murrelets has prevented convincing demonstration of population declines. However, populations have disappeared from the Lower Mainland and parts of the Strait of Georgia. Populations appear to be smaller in the region of Barkley and Clayoquot sounds where large oil spills, by-catch in gill nets, and logging of old-growth forests have all had potential impact. In the CIT region, the stresses on the population of murrelets appear to have been much less intense. There is no indication of decline but the population has never been studied in detail and its scale is largely a matter of conjecture. However, there has been a recent attempt to establish a practical value for the population baseline by using radar to count birds moving into specific watersheds.

The weak link between the exploitation of natural resources and variation in murrelet populations may have contributed to the bird’s long-term status as a threatened species. However, the biology of the bird does not lend itself to rigorous monitoring. It appears to produce only one egg per year and suffers a fairly high predation rate but the factors driving fluctuations in natural population, if there are any, are poorly understood. As a result, changes in population status can only be measured across decades and the accurate measurement of demographic parameters is problematic. However it is clear that the murrelet will be slow to recover from any population decline that does occur.

The Marbled Murrelet was selected as a focal species for the CIT ESA because of its apparent dependence on old-growth forest and the potential negative effects of exploitation of such forest. Our general understanding of the murrelet’s terrestrial habitat requirements suggest that its regional needs could be reflected by a simple model based on forest cover and other types of large scale mapping.

2.2.3.2 Methods

The model is based on one developed by the Marbled Murrelet Recovery Team for the Conservation Assessment (Part B) and IWMS species account (Marbled Murrelet Recovery Team 2003). We ran two versions of the model with Run 1 being the “most likely” and Run 2 being the “most + moderately likely” to capture suitable habitat for nesting Marbled Murrelets (see Table 2.6). Restrictions on elevation, age class, height class, and canopy closure class were relaxed in Run 2. Only Run 1 was used for the SITES analysis.

An amalgamated forest cover data set (1:20,000) created by BC MSRM, Resource Information Section from data submitted by tree farm licence holders (TFLs) and the timber supply areas (TSAs) was used to collect information on age class, height class, and canopy closure class for the murrelet nesting model. For Gwaii Haanas National Park Reserve, which was missing forest cover data, we used a GIS coverage created by the Gowgaia Institute depicting an updated logging history. The canopy closure class attribute, critical for running the nesting model, was missing from the forest cover data for TFL 39 on Haida Gwaii/QCI and TFL 39 on the Central Coast. A solution that correlated height and age classes to canopy closure class was developed that allowed us to run the model. Elevation from sea level was calculated from the TRIM Digital Elevation Model (DEM). We created a coverage in 100 m bands and built polygons for the ranges 100–900 m and >900 m. Due to the large size of the data set, for computing efficiency we split the



Page 51

TRIM DEM (1:20,000) into 350 separate segments, ran them, and then merged them back together to calculate distance from saltwater.

Table 2.6 Model structure for Marbled Murrelet habitats in the CIT study area

Run 1 (“Most likely”) Run 2 (“Most + Moderately Likely”)

Suitability class A B A B Distance from salt water 0–30 km 31–50 km 0–30 km 31–50 km Elevation: Central and North Coast Haida Gwaii/QCI

0–600 m 0–500 m

0–600 m 0–500 m

0–900 m 0–800 m

0–900 m 0–800 m

Stand age class 8 & 9 (>140 y) 8 & 9 8 & 9 (>140 y) 8 & 9 Stand height class 4+ (>28.5 m) 4+ 3+ (>19.5 m) 3+ Canopy closure class 4, 5, & 6 4, 5, & 6 3–7 3-7


The model showed a widespread and generalized distribution of murrelet habitat without any strong concentrations or significant isolated pockets (Map 8). As might be predicted from the model criteria, there was a strong tendency to include all lowland areas and valley bottoms. Sites for known nests were captured in the Queen Charlotte Islands but many of the nest sites around Mussel Inlet occurred at too high an elevation to be captured.

The model also suggested the presence of murrelet nesting habitat on small or low-lying islands and at various exposed localities along the coast. It is extremely unlikely that such sites are used by murrelets because of heightened levels of predation and the risk of storms destroying the nest. However, the total amount of habitat represented by such sites is insignificant. Their existence does suggest the possible inclusion of other unlikely sites further inland and the model should not be used at a finer scale than 1:20,000. The areas identified by the model are consistent with a broad band of opinion on what constitutes murrelet habitat at such a coarse scale.

The model indicated the presence of 216,093 ha of the most likely habitat on the Queen Charlotte Islands. No areas were added by relaxing the criteria (i.e., Run 2). There were 359,699 ha on the Central Mainland Coast and 154,736 ha on the Northern Mainland Coast that increased by 16,356 ha and 2,858 ha, respectively, when the criteria were relaxed.

2.2.3.3.1 Model applications and limitations

The model provides a general picture of the distribution of murrelet habitat along the coast of BC but it does not make any statements about variations in quality or usage. It may be that murrelets are fairly evenly distributed throughout their range and numbers vary in proportion to the amount of nesting habitat available but it is not very likely. At-sea observations suggest that there are a few areas (4 or 5) with regular concentrations of very large numbers of birds. These areas are hundreds of kilometres apart and travel between them is beyond the murrelet’s daily commuting distance. It seems likely that those marine concentrations could form the core of higher density nesting areas. This can only be tested by widespread comparison of radar counts and the location of nests by radio telemetry. Refer to Appendix 1.1.2.3.1 for a summary of model peer review comments.



Page 52

2.2.4 Focal Species: Northern Goshawk (Accipiter gentilis) Nesting Area and Foraging Area Habitat Suitability Models


The Northern Goshawk (Accipiter gentilis; goshawk) is a broad-winged, long-tailed raptor of medium size adapted to forest habitats where its quick acceleration and adept maneuverability make it a highly effective direct pursuit hunter (Mahon et al. 2003). The Northern goshawk was chosen as a focal species for the CIT ESA because of its strong association with mature/old-growth coniferous forests for foraging and nesting, and the possible impacts to this habitat from forestry activities.

Goshawks are found throughout circumpolar regions, and in temperate and boreal forests (Mahon et al. 2003). Within the CIT study area there are two subspecies of Northern Goshawk: the larger Accipiter gentilis atricapillus, which is found on the mainland, and the smaller provincially red-listed Accipiter gentilis laingi, found on Vancouver Island and Haida Gwaii/Queen Charlotte Islands (Campbell et al. 1990), and also possibly along the Central Coast (Cooper and Stevens 2000).

Goshawks prey on a variety of birds and small mammals - including Northwestern Crows (Corvus caurinus) and Marbled Murrelets (Brachyramphus marmoratus) on the coast Goshawks typically nest in mature/old-growth coniferous stands that are even-aged and have a closed canopy and open understory (Mahon et al. 2003). Fragmented forests with excessive openings favour other raptors adapted to more open habitats, some of which are predators of the goshawk (Utzig and Gaines 1998).

The Committee on the Status of Endangered Wildlife in Canada (COSEWIC) has designated the Northern Goshawk (Accipiter gentilis) as “Not At Risk” and the Queen Charlotte goshawk (Accipiter gentilis laingi) was uplisted in November 2000 to “Threatened” status. In B.C., Accipiter gentilis laingi is Red-listed, and the Alaska Natural Heritage Program lists it as a “Species of Special Concern. “

2.2.4.2 Methods

A goshawk breeding territory has three hierarchical components: nest area, post-fledging area and foraging area (Reynolds et al. 1992 in Mahon et al. 2003). Habitat Suitability Index (HSI) models for the Nesting Area and Foraging Area components of a goshawk territory were developed for the 3 Land and Resource Management Planning (LRMP) areas in the CIT ESA study area (North Coast, Central Coast, Haida Gwaii/Queen Charlotte Islands). The models are based on those developed by Mahon et al. (2003) for the North Coast Forest District. Model structures were based on the Habitat Suitability Index (HSI) methodology developed by the U.S. Fish and Wildlife Service (1981). In developing their models, Mahon et al. (2003) combined expert opinion, observed habitat characteristics at goshawk nest areas in the Coastal Western Hemlock biogeoclimatic zone, and relevant literature.

Goshawks consistently select key structural attributes and forest species composition for nesting habitat (for a review of these attributes, see Mahon et al. 2003). Table 2.7 lists these attributes.



Page 53

Table 2.7 Key structural attributes and forest species composition of goshawk nesting habitat (Mahon et al. 2003)

Key structural attributes Supporting literature

Mature/old-growth stand structure

Relatively closed canopies with corresponding open understories.

Kennedy 1988; Hayward and Escano 1989; Reynolds et al. 1992; Mahon et al. 2003

Forest species composition Supporting literature

Dominated by western hemlock

≥ Age class 8 (141 years)

≥ Height class 4 (28.5 m)

≥ Canopy closure class 5 (56%)

Mahon et al. 2003; Mahon and Doyle 1999; Schaffer et al. 1999.

For the CIT ESA goshawk models, parameters and values for the North Coast models were taken from the Mahon et al. (2003) model which covered the same area, while those for the Central Coast and Haida Gwaii/QCI models were developed with the advice of Frank Doyle, one of the developers of the North Coast Forest District model, and Erica McClaren, a Species At Risk biologist with the BC Ministry of Water, Land and Air Protection. The habitat ratings resulting from these models represent relative values suitable for comparisons across the CIT ESA study area.

2.2.4.2.1 Nesting Area HSI model

The Nest Area HSI model follows a limiting factor, non-compensatory approach (Mahon et al. 2003). Table 2.8 provides a description of nesting area suitability indicators.

Table 2.8 Indicators of nest area suitability (Mahon et al. 2003)

Indicator Description

Height class The structural maturity of a stand, and trees within a stand, form the fundamental basis for nesting suitability for goshawks. Individual trees must have large enough branches to support the nest structure. Projected stand height class was used as a surrogate for structural stage in the model.

Canopy closure

After the fundamental requirement of a “mature” forest stage, canopy closure is the next most important structural variable relating to nest area suitability.

Tree species Suitability depends on the form and structure of the trees, the stands, and varies with site and age. Overall stand forest type suitability ratings are calculated by multiplying the species rating by its percentage composition and summing the individual species ratings for all types in the stand.

Edges Data indicates that goshawks tend to avoid locating nests near forest edges (Mahon et al. 2003). Avoidance is strongest 0–50 m from a “hard” edge. Hard edges occur where mature forest meets non-forested or early seral habitats and the difference in height was >10 m. Hard edges occur around regenerating cutblocks, roads, human settlement/development, swamps, swamp forest, wetlands, brush patches, lakes, rivers, and ocean.



Page 54

The equation used to calculate the nesting area suitability ratings is:

Nest area suitability = height class rating x canopy closure rating x tree species rating x edge rating

Parameters and values used for the suitability ratings are detailed in Appendix 1.1.2.1

2.2.4.2.2 Foraging Area HSI model

The foraging area is the entire area used by the adults. Numerous studies conducted in coastal B.C. and southeast Alaska found that prey associated with mature forests such as red squirrels, forest grouse, and passerines dominate goshawk diets.3 This favours hunting primarily in mature/old-growth forest areas with high canopy closure, and a clear understory, a habitat that allows goshawks to move freely under the canopy, allows good visibility of its prey and also provides ample perches from which it hunts (Squires and Reynolds 1997 in Mahon et al. 2003). The Foraging Area HSI model follows an additive, compensatory approach (Mahon et al. 2003).

The equation used to calculate the foraging area suitability ratings is the higher value of:

Foraging area suitability = height class rating + age class rating+ canopy closure rating + elevation rating + slope rating + shoreline rating

Or

Foraging area suitability = non-productive/non-forested habitat rating

Table 2.9 provides a description of foraging area suitability indicators. Appendix 1.1.2.1 provides full details of parameters and values for the suitability ratings.

Table 2.9 Indicators of foraging area suitability (Mahon et al. 2003)


Height class Mature/old-growth habitats are the primary foraging habitats for goshawks. Projected height class was used to capture this indicator.

Age class Projected age class was used as a correlate to structural stage. Including both age and height class, recognizes their strong correlation and the weighting they influence on the model. Similar to height class, age class is assumed to approximate structural stage.

Canopy closure class Moderate to high canopy closure tends to correlate to open understories, which goshawks use as flyways while hunting.

Elevation

Empirical telemetry data from Iverson et al. (1996) indicates that goshawks foraged less at higher elevations.

Slope

Lower gradient slopes are given a higher rating, as these are typically richer sites producing larger trees and are associated with higher prey densities. Similar to elevation this factor is weighted lightly in the model.

3 Mahon et al. (2003) refer to Doyle and Mahon 2001, Mahon and Doyle 2001, McClaren 2001, Chytyk and Dhanwant 1999, Iverson et al. 1996, Lewis 2001, Austin 1993, Doyle and Smith 1994, Beier and Drennan 1997, Squires and Reynolds 1997, Good 1998, Stephens 2001



Page 55


Shoreline Headlands and fiords heavily bisect the CIT study area and as a result much of the area is relatively close to the shore. Goshawks appear to use the forest edge as cover as they hunt the rich diversity of prey (northwestern crow, alcids, gulls, ducks) that are available along the shoreline, and which are common prey of goshawks in these areas (Lewis 2001). Areas < 300 m from shore receive a higher habitat rating.

Non-productive & Non-forested habitats

Many non-forested habitats occur in the CIT study area that may be used by goshawks to forage.

2.2.4.2.3 GIS and data sets

A GIS (ESRI ArcGIS and ArcView) was used to run the algorithms and spatially display the outputs of the models. The models used a 1:20,000 TRIM DEM to calculate distance to shoreline, elevation, and slope; an amalgamated 1:20,000 forest cover data set composed of data from tree farm license holders (TFLs) and timber supply areas (TSAs) in the CIT study area, to calculate height class, canopy closure class, tree species, an edge rating, and non-productive/non-forested habitats ratings. Tree species were added and ranked for the Central Coast that were not used in the North Coast or Haida Gwaii/QCI models. Also, because a portion of the study area did not have canopy closure class in the forest cover data set (TFL 39 on Haida Gwaii/QCI and TFL 39 on the Central Coast), theoretical canopy closure class was calculated based on height and age classes. The assumption was that if a tree was a certain height and age range, this correlated with its canopy closure range.


Nesting and foraging habitats from the above models are combined and displayed in Map 9. The result of running the algorithms and GIS data was a grid with Habitat Suitability Index ratings ranging from 1 to 3, with 1 being High, 2 being Moderate, and 3 being Low Suitability nesting and foraging habitat.

2.2.4.4 Model Applications and Limitations

The Nest Area model should provide an accurate prediction of suitability at the stand level, but cannot predict probability of use (Mahon et al. 2003). The Foraging Area model is less sensitive to behavioral isolation of habitats and represents a rough index of probability of use by goshawks. However, Mahon et al. (2003) noted that overall confidence in the Foraging model is significantly lower than for the Nest Area model for several reasons. First, the Foraging Area model is a combined, multi-species prey abundance + prey availability model + goshawk use model. Therefore, it is a very general model that provides the best components and average component ratings for several factors, but loses specificity and accuracy with respect to any individual factor. Second, no local empirical data was available to support the assumptions built into the model. Third, prey abundance, prey composition, and prey availability are known to vary considerably over time (Mahon et al. 2003). Refer to Appendix 1.1.2.3.2 for a summary of model peer review comments.



Page 56

2.2.5 Focal Species: Black-Tailed Deer (Odocoileus hemionus sitkensis and O. hemionus columbianus) Winter Range Model


The black-tailed deer (Odocoileus hemionus sitkensis and Odocoileus hemionus columbianus) was chosen as a focal species for the Central Coast and North Coast portions of the study area because losses of old-growth forest cover has a high potential to negatively affect deer populations (Suring et al. 1992). The value of old-growth forest ecosystems to black-tailed deer is directly related to the composition, structure, and productivity of these forests (Harestad 1985). The complex canopy structure of old-growth forests allows sufficient light to penetrate, promoting the growth of a diverse set of vascular and non-vascular plants for forage; as well as providing for interception of snow.

Deer abundances also ultimately affect predator-prey relationships. Along the BC coast, black-tailed deer make up a large portion of gray wolves’ diets during the summer and fall months (Darimont and Paquet 2000; McCrory et al. 2003). As researchers have noted, conservation of wolves requires the conservation of prey species and their habitats (Darimont and Paquet 2000, McCrory et al. 2003). We targeted black-tailed deer winter range as a key component of their habitat, which acts as a limiting factor in deer abundance (see Suring et al. 1992; Darimont et al. 2003, McCrory et al. 2003). The amount and duration of snowfall an area receives strongly influences its ability to support deer (Suring et al. 1992). For example, Mech et al. (1971) found that snow accumulations during single and consecutive winters were directly related to fawn-doe ratios and annual changes in deer populations. Other studies have also reported a correlation between snow depth and deer mortality or population levels (Severinghaus 1947; Edwards 1956 in Suring et al. 1992). Jones and Bunnell (1984) found that on northern Vancouver Island, severe winters resulted in reduced diversity of forage available to deer, restricted mobility of deer, caused mortality of deer, and lowered recruitment rates of deer.

During periods of deep and prolonged snow, deer look for old, high-volume forests on gentle to moderate slopes at low elevations (Suring et al. 1992; McCrory et al. 2003)—likely because this forest type is most effective at intercepting snowfall (Kirchhoff and Schoen 1987). Old-growth forests also provide ample winter forage, such as nutrient-rich shrubs, lichens, and litterfall (Kirchhoff and Schoen 1987). Studies in southeast Alaska have found that with increasing snowfall, deer decrease their use of open habitats such as muskegs and young clearcuts, and increase their use of old-growth forests (Kirchhoff and Schoen 1987; Schoen and Kirchhoff 1990).

2.2.5.2 Methods

The CIT ESA black-tailed deer winter range model is based on an “old-growth, deer winter-range” model developed by McCrory et al. (2003). We applied a GIS to spatially identify deer winter range in the study area. Black-tailed deer were used as a focal species for the North and Central Coast portions of the study area. The study team decided not to model deer winter range on Haida Gwaii/QCI, due to the negative impacts associated with deer introductions to the islands (Sharpe et al. 2002).

The model included the following geographic data sets: forest cover data to identify forest type, age class, and volume class; and a digital elevation model (DEM) generated from 1:20,000 TRIM



Page 57

maps to calculate elevation and slope steepness. Table 2.10 details the model parameters and values.

Table 2.10 Black-tailed deer winter range model parameters and values (McCrory et al. 2003)

Indicator Value

Elevation Between sea level and 500 m

Forest type Western hemlock (Tsuga heterophylla) dominated forest types.

Age class 121–140 years to >500 years

Volume class Low (151–200 m3/ha) to High (850 m3/ha)

Slope steepness < 400


The result of the GIS modelling is a map of winter range displayed as a binary grid; cells with a value of 1 are identified as winter range. The model identifies 688,775 ha of black-tailed deer winter range in the study area (517,835 ha are in the Central Coast, and 170,628 ha are in the North Coast). Habitat was distributed evenly throughout the 2 LRMP areas and as expected, in areas featuring older coastal western hemlock forests including many of the islands.

The model and maps were reviewed by experts in B.C. and southeast Alaska. Review comments were consistent in two key areas: 1) parameter values were too general in scope; and 2) key indicators were missing from the model that could result in the model identifying the wrong areas for deer winter range habitat.

McCrory et al. (2003) omitted aspect to “reflect a wide range of winter conditions including north slopes that might be used during mild winters” (p. 50). Darimont et al. (2003) in their subsequent application of the model on the Central Coast, hypothesized that aspect was highly correlated with forest type and “in steep terrain such as the topographically complex mainland of Heiltsuk Territory, aspect may not be a good predictor of sun exposure given the low angle of winter sun at this latitude” (p. 6). In both reports, elevation and characteristics of the forest canopy were identified as more important variables to include in the model. However, aspect is considered an important indicator of deer winter ranges (K. Brunt, pers. comm.; M. Kirchhoff, pers. comm.). The assumption that aspect was highly correlated with forest type is questionable on the grounds that variables not measured on an interval-level scale (e.g., forest types, names) cannot be statistically correlated (M. Kirchhoff, pers. comm.). From an ecological viewpoint, it is doubtful that hemlock dominated forests (CWH) occur on just one aspect (or 2 or 3). Species composition is more a function of soil characteristics than aspect. For example, cedar tends to thrive on poorly drained soils, spruce dominates on alluvial and colluvial soils, and hemlock dominates on organic soils. A more even-aged stand on wind-prone southern aspects (structural differences) might be expected, but not necessarily different species (compositional differences) (M. Kirchhoff, pers. comm.).

Of the 5 model variables, including Slope Steepness has weak empirical support in the literature. For example, a steep south-facing slope that has good forest cover may have quite low-snow



Page 58

depths and good light penetration and understory cover (M. Kirchhoff, pers. comm.). Another indicator that was not included in the model is exposure. Good exposure during the winter occurs on steep, south aspect sites that are unshaded by adjacent mountains (K. Brunt pers. comm.).

With modifications to the model (e.g., the inclusion of aspect, possibly exposure, and the removal of Slope Steepness), it would more accurately reflect black tailed deer winter range in the CIT study area. McCrory et al. (2003) and Darimont et al. (2003) applied this model on small portions of the CIT study area, which may be the appropriate scale and use for it. However, due to time and capacity constraints, the ESA focal species team was not able to modify and rerun the model based on expert review recommendations. As a result, the model was not used as an input to the SITES/MARXAN model. Refer to Appendix 1.1.2.3.3 for a summary of model peer review comments.

2.2.6 Focal Species: Mountain Goat (Oreamnos americanus) Winter Habitat Model


The genus Oreamnos is represented by a single extant species, O. americanus, or the mountain goat, found only on the North American continent. The mountain goat occupies steep, rugged terrain in the mountains of northwestern North America, with native populations in British Columbia, the northern Cascades of Washington and the northern Rocky Mountains of Montana and Idaho. British Columbia is home to up to 60% of the continental mountain goat population (cited in Wigal and Coggins 1982).

Generally, mountain goats inhabit alpine and subalpine habitats in all of the mountain ranges of British Columbia. The characteristically rugged terrain is comprised of cliffs, ledges, projecting pinnacles, and talus slopes. In particular, it has been noted that the availability of winter range may be limiting for many mountain goat populations (Fox and Smith 1988; Poole and Mowat 1997; Sims 1999). Winter habitats may be low elevation habitats where snow accumulation is low, or high elevation habitats where wind, sun or precipitous terrain adequately shed snow from foraging habitats. In coastal BC areas, mountain goats generally move to south-facing, forested areas that offer reduced snow loading and increased access to foraging. These winter habitats are limiting, and typically goats show high site-fidelity to these areas (Pollard 2002a). Thus, direct (e.g., timber harvest) or indirect impacts (e.g., increased disturbance or road access) to these wintering habitats have can have large impacts on the local and regional mountain goat populations (Fox et al. 1989).

Declines of mountain goat population numbers and productivity have been documented in many regions across their range. Most of these declines have been linked to the noncompensatory response of goat populations to increased hunter harvest mortality and population reduction (reviewed in Peek 2000). Recovery of mountain goat populations can often be prolonged, and is often confounded by population fluctuations in response to adverse weather conditions (Peek 2000). Increased access (e.g., roads) is linked to increased hunting pressure in many harvested species, including mountain goat, and, as resource extraction activities increase access, near-by goat populations may become increasingly susceptible to over-harvest.

The mountain goat was selected as a focal species for the CIT spatial analysis because of the sensitivity of this species to direct impacts of forest cover removal from in limited winter ranges,



Page 59

as well as the potential direct and indirect mortality associated with increased access. In many regions of the CIT study area, past surveys and fine-scale habitat modelling has identified present populations and associated winter habitats (see Pollard 2002b). These data and the modelling results were used to identify the highest priority mountain goat areas in the CIT ESA. Additionally, a coarse-scale, spatially explicit model was developed to predict potential mountain goat winter habitats in areas not included by these previous efforts, and also to predict potential future habitat potential (i.e., habitat capability).

2.2.6.2 Methods 2.2.6.2.1 Mountain goat winter habitat identification

The characteristics of winter habitat vary across goat range, likely due to the variable conditions that may cause snow to be removed (summarized in Peek 2000; Pollard 2002b). Slopes are typically greater than 65% in most studies. The habitats used by goats for foraging are highly variable, but are typically found within 400 m of security terrain, and include many types of open habitats. In coastal regions of BC and Alaska, the use of mature, closed canopy forest adjacent to security habitat provides forage and shelter areas for goats in the winter (Suring et al. 1988; Pollard 2002b). Mountain goats typically move between winter habitat and summer habitat areas, and, while these seasonal migrations may be up 24 km in some regions, such seasonal movements have been found to be limited to short distances in coastal regions (see review in Pollard 2002b). Thus, in coastal areas, suitable winter habitat is limited to those areas within close proximity to potential summer habitats (Pollard, pers. comm.).

2.2.6.2.2 Existing data and winter habitat models

As previously mentioned, there has been substantial effort previously invested across much of the study area to identify and map existing goat populations and associated winter habitats. These data were provided by MWLAP and, for the North Coast region, by the LRMP efforts (Pollard 2002a, 2002b). All of these efforts represent a combination of GIS-based modelling of potential winter habitat subsequently modified using aerial photo interpretations and known goat locations. “Mountain goat winter range was to be identified using a two tiered approach considering known and suspected patterns in sub-regional and stand-level winter habitat selection. South facing, steep slopes with elevational connectivity to summer ranges were identified at the sub-regional scale on 1:50,000 mapping. Areas with a high probability of containing winter range were then examined in detail using aerial photographs and 1:20,000 TRIM maps to identify escape terrain. Historical, anecdotal and concurrent aerial survey information was used to identify areas with known winter mountain goat use” (Pollard 2002b, Executive Summary). The habitat attributes and assumptions used in the initial GIS habitat model are summarized in Table 2.11, and form the basis for the potential winter habitat model developed for the extent of the CIT area (described below). A complete description of the methods used in the North Coast is provided in Pollard (2002b). These methods represent approaches taken across the data sets provided by MWLAP for the CIT ESA (Tony Hamilton, pers. comm.).



Page 60

Table 2.11 Landscape features and assumptions used in 1:50,000 mapping for mountain goat winter range in the North Coast region of the study area (Pollard 2002a)

Landscape feature Mapping assumptions Supporting literature 4

Slope Mountain goats winter on slopes between 30o and 65o.

*1, *2, *4, *6

Aspects Mountain goats winter on aspects between 135o and 250o.

*2, *3, *4, *5, *6, *7,

Vegetation Mountain goats winter in or beside coniferous stands older than 60 years.

*6, *7, *8, *9, *10, *11,

Accessibility Mountain goat population movement is limited by extensive icefields or water.

*12, *13

Adjacent summer range Mountain goat winter habitat is connected elevationally to suitable summer range.

*2, *14, *15

2.2.6.2.3 Potential mountain goat winter habitat model

We have conducted additional modelling to provide predictions of potential winter goat habitat across the CIT area. This effort aims at “filling in the gaps” in the spatial extent of the existing efforts, as well as identifying potential, but unoccupied habitats. This effort builds on the foundational GIS-based modelling used by Pollard (2002b) to identify areas for finer-scale habitat identification (summarized in Table 2.11). Additionally we have drawn upon other existing goat winter habitat modelling done in other similar regions (e.g., southeast Alaska). The potential habitats predicted by this model receive lower consideration in the CIT ESA analyses than the previous efforts that included finer-scale habitat identification based on aerial photography interpretations and existing goat survey data. We used the 50 m resolution DEM to identify potential escape terrain as warm aspect, steep habitats adjacent to potential foraging habitats, as well as near potential summer habitat. See Table 2.12 for habitat attributes and assumptions.

Table 2.12 CIT area-wide mountain goat potential winter model definitions and assumptions


Slope (winter escape terrain) Mountain goats winter on slopes between 30o and 100o.

See Table 2.11;

*1, *2, *3, *4, *5, * 6

Aspects (winter escape terrain) Mountain goats winter on aspects between 135o and 250o.

See Table 2.11;

*3, *7

4 *1 Schoen & Kirchoff (1982); *2 Foster (1982); *3 Hebert & Turnbull (1977); *4 Smith (1977); *5 Fox (1977); *6 Fox et al. 1989; *7 Smith (1994); *8 Hjeljord (1973); *9 Schoen et al. (1980); *10 Smith (1986); *11 Fox & Smith (1988); *12 Hazelwood, pers. comm.; *13 Banfield (1974); *14 Russell (1974); *15 Chadwick (1977) [*as cited in Pollard 2002b]. 5 1 Sims (1999); 2 Poole & Mowat (1997); 3 Fox et al. (1989), Lowell et al. (1988), Smith (1986); 4 Joslin (1986), as cited in Peek (2000); 5 Johnson (1983), as cited in Peek (2000); 6 Haynes (1992); 7 Adams & Bailey (2002), Gross et al. (2002).



Page 61


Winter forage Forested habitat (all seral stages) within 400 m of escape terrain; in regions of the study area >50 km from the ocean may not be limited by snowfall and so also included grassy or shrubby habitat including alpine.

See Table 2.11;

*1, *2, *3, *5;

Reviewer comments

Elevation 600–1200 m Reviewer comments

Accessibility Mountain goat habitat excludes extensive icefields or water.

See Table 2.11;

Reviewer comments

Adjacent summer range Mountain goat winter habitat is within 3 km of suitable summer range (identified as open, shrubby or grassy habitats including alpine areas)

See Table 2.11;

B. Pollard, pers. comm.;

Reviewer comments

Habitat patch size and adjacency Removed escape terrain <1 ha in size and>400 m from adjacent escape terrain

B. Pollard, pers. comm.

2.2.6.2.4 Input into the CIT ESA

The existing MWLAP data and fine-scale modelling on occupied or highly probably winter habitats provides the best information on current goat distribution and occupied or high quality winter habitats. These spatial data were combined with the CIT-wide modelling which provides identification of areas without previous data, as well as predictions about future potential winter habitats. The areas identified in the MWLAP and North Coast LRMP data were given a 3-fold weighting relative to the potential habitat modelling completed study area wide. This weighting reflects both the scale and the confidence in the modelling and data, while still allowing the ESA to consider those areas where there has not been efforts to map existing goats populations, as well as to consider potential future goat winter habitats across the study area.


The combined MWLAP and North Coast LRMP data and modelling identified approximately 300,000 ha of occupied winter goat habitat in the CIT study area. These habitats are not equally distributed across the study area, due to the project-specific nature of each effort that contributed these data. The CIT-wide potential winter habitat model overlaps 65% of these occupied winter mountain goat habitats, and predicts an additional approximate 1 million ha of present or future potential winter habitat across the study area (Map 10). This represents 2.3% of the mainland portion of the study area.

As mentioned earlier, the habitats that were identified in the data and modelling provided by MWLAP and the North Coast LRMP represent finer-scale habitat identification efforts and incorporate known goat locations. The areas identified in these efforts were given three-fold higher weighting than those habitats identified in the coarser-scale CIT-wide potential habitat model. The CIT-wide model does not include the finer-scale aerial photo-interpretation, and is based primarily on the model attributes used in the North Coast winter goat habitat identification. It identifies habitats with potential values for present or future goat populations,



Page 62

but unknown current occupation. This weighting reflects both the scale and the confidence in the modelling and data, while still allowing the ESA to consider those areas where there has not been efforts to map existing goats populations, as well as to consider potential future goat winter habitats across the study area. Existing winter goat habitat identification, provided by MWLAP and the North Coast LRMP effort and predicted potential goat winter habitat in the CIT region are displayed in Map 10. Refer to Appendix 1.1.2.3.4 for a summary of model peer review comments.

2.2.7 Focal Species: Grizzly Bear (Ursus arctos) Habitat Suitability and Capability Models


Grizzly bears (Ursus arctos) are found throughout coastal British Columbia, with the exception of the Georgia Depression ecoprovince, Vancouver Island, Queen Charlotte Islands, and the Coastal Douglas-fir (CDF) biogeoclimatic zone. Grizzly habitats range from sea level and estuaries, up into subalpine areas. Coastal grizzly bears are mostly solitary, intra-specifically aggressive omnivores that typically have large seasonal and annual home ranges (RIC 1999).

Coastal grizzlies require habitat that provides for their nutritional, security, thermal, reproductive and “space” needs. To meet these varied needs, bears use an array of habitats, ranging from subalpine to valley bottom, old growth to young forest, and wetlands to dry areas. With the exception of denning areas and avalanche chutes, the prime habitat of coastal grizzlies occurs predominantly below treeline and is largely concentrated in valley-bottom ecosystems often associated with important salmon streams.

Coastal grizzly bears are opportunistic, feeding on early green vegetation such as sedges (Carex spp.) and skunk cabbage (Lysichiton americanum) in the estuaries and rich wetlands and swamps. As the season progresses, they often use avalanche chutes where they again feed on a variety of food sources, including cow-parsnip (Heracleum lanatum). As berries ripen, they feed on floodplains and side hills where devil’s club, salmonberry, raspberry, black twinberry, elderberry, and a variety of blueberries are found. In the late summer and fall, coastal grizzlies focus on spawning salmon (Oncorhynchus spp.) often traveling long distances to fish.

Key components of coastal grizzly bear conservation include: 1) managing motorized vehicle and aircraft access to minimize bear displacement and mortality; 2) maintaining habitat quality and quantity at multiple scales, including landscape-level forage supply and habitats at the stand level; 3) minimizing potential for bear displacement and habituation as a result of human activities such as wildlife viewing; 4) regulation of hunting levels and providing benchmark areas where hunting of bears is not permitted; and 5) minimizing potential for bear-human interaction by promoting the use of “bear awareness.”

Grizzly bears were chosen as focal species for the CIT ESA because they, like other large carnivores, can be keystones (transporting salmon away from spawning channels), indicators (because they are susceptible to a wide variety of human influences and have low population densities), and umbrellas (representing a number of species because of their use of such a wide variety of habitats). Grizzly bears were a main focal species in a similar conservation assessment in the Greater Yellowstone Ecosystem (Noss et al. 2002).



Page 63

The CIT ESA grizzly bear model is a developmental extension of the provincial grizzly bear estimation process commonly referred to as the Fuhr-Demarchi method (Fuhr and Demarchi 1990; Hamilton and Austin 2002). The premise behind the approach is that different ecological units vary in their ability to support grizzly bear food resources and that such variations are linked (linearly) to bear density. The higher any one land area is ranked for its ability to provide grizzly bear foods, the higher the density estimator attached to it. Each of the 6 ranks assigned to a habitat represent a density range (minimum and maximum) from zero to a maximum of 100 bears/1000 km2. Given the potential consequences of overestimating grizzly populations, the Ministry of Water, Land and Air Protection manages populations on the basis of the minimum estimates (Hamilton and Austin 2002). Although this model was originally linked to bear density, reviewers of the draft output recommended breaking that link. Instead of attempting to translate habitat effectiveness into bear density, the model simply reports grizzly bear habitat effectiveness across the region.

The subjectivity of Fuhr-Demarchi has been reduced by applying knowledge from several British Columbia coastal grizzly bear research projects (Hamilton et al. 1986; Hamilton 1987; MacHutchon et al. 1993; Himmer and Gallagher 1996), a DNA/Hair Mark-Recapture population estimate for the Kingcome and Wakeman watersheds (Boulanger and Himmer 2001) and ongoing population monitoring in the Owikeno Lake basin (Himmer and Boulanger 2003) and along the north side of the lower Nass River (Demarchi 2003). Additional information for the Northwest part of the study area near Atlin was obtained from a small grizzly bear research project associated with the proposed Tulsequah Chief Mine and Access Road (Wellwood and Diemert, pers. comm.).

2.2.7.2 Methods

The grizzly bear model used the following data as indicators of habitat suitability:

• Broad Ecosystem Units (BEU)

• TRIM 1:20,000 Digital Elevation Model (DEM)

• Salmon biomass estimates

• Roads/road density

Data assembly and processing covered a larger area than the CIT study area to apply a broader “coastal” grizzly bear context, to ensure consistency with previous grizzly bear mapping, and to cover a larger coastal planning project area. Grizzly bears were modelled in their occupied range6 only and results were presented for both full and partial Grizzly Bear Population Units (Hamilton and Austin 2002).

The current application of Fuhr-Demarchi is at the ecosection/variant/phase level of British Columbia’s biogeoclimatic ecosystem/ecoregional classification (Meidinger and Pojar 1991). The

6 “Occupied range” is the term used by British Columbia’s Grizzly Bear Conservation Strategy to describe the land area inhabitated, on an annual basis, by overlapping resident adult female grizzly bears. Occasional transient bears may be sighted outside of occupied range. Some coastal islands have periodic sightings of individual animals or, on rare occasions, adult females with cubs. However, these areas are not considered occupied for the purpose of estimating populations or setting habitat and population objectives.



Page 64

CIT ESA grizzly bear model increases a level in resolution and spatial accuracy to Broad Ecosystem Units (BEUs). BEUs subdivide Variants and Phases into a permanent area of the landscape that supports a distinct type of dominant vegetation cover, or distinct non-vegetated cover (such as lakes or rock out-crops) (RIC 2000). One of the clear advantages of using the Broad Ecosystem Inventory (BEI) for modelling is that each forested unit is defined as including potential (climax) vegetation and associated successional stages. Unit classification is further subdivided by “modifiers” (RIC 2000) that characterize topographic and soil features (e.g., coarse soils, north facing aspects). These enhancements enable a more rigorous distinction between habitat capability (idealized value) and habitat suitability (current value independent of human displacement) than under Fuhr-Demarchi (1990). Individual unique combinations of ecosection, biogeoclimatic zone, subzone, variant, phase, BEU, modifier, and seral (successional) stage were rated in the standard 6-class system for capability and suitability to support grizzly bears (RIC 1999). Care was taken to rate all seral stages that could occur, rather than just those that do occur in the map base, to ensure capability was determined correctly. Rating assumptions by unit are described in Appendix 1.1.2.2. Mid-seral (coniferous) stages of forested BEUs were typically ranked low because few bear forage species survive the low light conditions under closed coniferous canopies (Michelfelder 2003).

A second major difference between the Fuhr-Demarchi procedure and this model is the explicit inclusion of estimates of Pacific salmon biomass as they influence bear density. Estimates of biomass were derived from Department of Fisheries and Oceans salmon escapement records7 linked to explicit watershed boundaries.8 A check of this link was completed for the large river systems that have more than one watershed along their main reach lengths (e.g., Nass, Skeena, Dean, Babine) by examining the spawning area location notes in the DFO salmon escapement catalogs. Spawners were assigned to any one watershed/reach length on a majority basis. Historic maximums (by species by run) were used to establish minimum and maximum capability. A ten-year average of recent escapements from 1993 to 2002 was used to establish the contribution to bear suitability. The model attempts to account for bear population depression caused by recent salmon declines by incorporating these declines in the 10 year suitability average (e.g., sockeye in the Owikeno Basin and pink salmon in Knight and Loughborough Inlets). Biomass estimates were derived from literature reports of average spawning weights by species and sex and assuming a 50/50 sex ratio.

The importance of spawning salmon to coastal bear density has long been recognized (e.g., Hilderbrand et al. 1996) but no study has yet examined density contributions from salmon separately from vegetative (or other animal) food sources. A “1/3” rule was assumed for this modelling exercise. That is, two thirds of the starting capability and suitability estimates were derived from interpreted BEI, one third from the dietary contribution of salmon. Catastrophic effects on cub production and survival that may result from severe salmon declines were not modelled.

7 Salmon escapement data were not available for rivers north of Stewart—the model likely underestimates current and potential densities in the lower reaches of the major Northwestern river valleys, e.g., Whiting, Unuk, Taku, Stikine. 8 Department of Fisheries and Oceans escapement records are limited by inconsistent stream monitoring effort and the difficulty of estimating escapement in turbid streams. Approximately 8% of the provincial total escapement did not link to any watershed boundary and were therefore excluded. Escapement records for several important northern rivers (e.g., Taku) were unavailable for this modelling exercise.



Page 65

Digital Elevation Model (DEM) information on slope has also been incorporated into both capability and suitability estimates. The rugged coastal topography often restricts highly suitable grizzly bear habitats to lower slopes and valley bottoms. This model limits the capability and suitability of steep slopes to recognize this restriction. Slope was assigned into 3 classes: 0–60%, 60–75%, and >75%. Steep slopes were not entirely eliminated from the density estimator because grizzly bears will use such slopes on occasion, particularly as they seek the emergent vegetation on avalanche tracks (MacHutchon et al. 1993). Instead, slopes from 60 to 75% were “stepped down” by 25% and slopes greater than 75% were stepped down by 50%. That is, for any one starting capability or suitability estimate, the model applies suitability reductions of 25% and 50% respectively for these two slope classes.

The fourth departure from the Fuhr-Demarchi procedure is the explicit modelling of the influence of roads on habitat displacement. Although a large number of kilometres of open road are found in coastal British Columbia, their displacement influence is mitigated because many of them are part of isolated road networks built for logging. Traffic levels during forest harvesting can be high on such networks, but low or nil during economic downturns, fire seasons or between logging phases or “passes” (Archibald et al. 1987). In addition, some bears may show high tolerance to moving vehicles and gradually habituate to logging traffic. Other bears may have little choice than to enter the zone of influence around the road, as coastal roads are often located near important seasonal habitats in river valley bottoms to minimize gradients and reduce road construction costs.

Isolated road networks may remain open long after active harvesting has been completed. Such roads may constitute a significant mortality risk well into the future. Rather than attempt to model displacement only on “active” road networks and separately model mortality risk from connected road networks vs. isolated ones, road density was used as a combined surrogate for both potential limiting factors. All mapped roads were considered to be open and to have traffic volumes that would displace bears from the immediate vicinity around them. This recognized overestimate of displacement was a deliberate attempt to somehow recognize the mortality risk potential from open coastal roads.

Habitat suitability was further reduced for various road densities in a manner similar to the application of “step downs” for slope. The most recent coastal road coverages were assembled for road density class assignment using a “moving window” approach (Mace et al. 1999). The reference density of 0.6 km/km2 (USFWS 1993) was reduced by 25%, 0.6 to 1.2 km/km2 by 50% and >1.2 km/km2 by 75%. Suitability was not reduced to zero, even for the highest road densities, because studies indicate that some grizzly bears will still use high value habitats adjacent to high levels of vehicle traffic (Yost and Wright 2001; Gibeau et al. 2002). However, studies also indicate that habituated bears are the most likely to come into conflict with humans and have an elevated mortality risk as a result (Benn 1998). Estimates resulting from this last step in the model are referred to as “effectiveness” estimates. That is, ineffective habitat may be entirely suitable, but limited in its use by the displacement influence of vehicles on roads. Effective habitat is both suitable and relatively free from any human influence.

The final major departure from Fuhr-Demarchi is the application of individual watersheds as units for summarizing and mapping estimated current effectiveness, potential capability, suitability, and effectiveness. Although the model assigns effectiveness to individual watersheds for comparative purposes, the approach is artificial. Most watersheds are smaller than average



Page 66

bear home ranges and actual distribution of grizzly bears varies significantly on a seasonal and annual basis. The model’s greatest utility is in comparing relative grizzly bear habitat priorities at the watershed level across the study area.


The primary model outputs are the GIS coverages and associated databases of current minimum effectiveness classes for each of the 5,983 modelled watersheds in occupied grizzly bear range within the study area (Map 11).

2.2.7.3.1 Model applications and limitations

Caution should be exercised in applying proposed conservation targets for partial GBPUs. Ecological characteristics and human activities outside the study area boundary in partial GBPUs influence grizzly bear distribution and abundance inside the modelled study area.

No attempt was made to model the potential displacement influence of aircraft. Regardless, there is some evidence that grizzly bears are displaced from landing and takeoff areas, particularly of helicopters, at or near important seasonal habitats. Nor does the model attempt to examine potentially increased mortality risk around population centres. Third, this model does not attempt to account for historic human-caused mortality on current population estimates. There is considerable debate on how to model such impacts appropriately and doing so objectively is one of the key recommendations of the recent Provincial Grizzly Bear Science Panel (Peek et al. 2003).9

Comparison of the current effectiveness with the proposed target effectiveness can be instructive: watersheds with high correspondence between current and proposed target effectiveness may be high priorities for protection or “light touch” ecosystem-based management. Watersheds with high discrepancy between current estimates and proposed target estimates may be good candidates for the recovery of suitability through silvicultural intervention or for prescribed fire in the coast-interior transition zone. In other such situations, effectiveness might be recovered through road deactivation or other, stronger, motorized access control methods such as bridge removal.

The model has been designed to address the full spectrum of ecosystem-based management of grizzly bears in coastal and northwestern British Columbia. It can assist in the identification of an appropriate reserve network with nodes of protected areas that may buffer and link across subregions. However, perhaps more importantly, the model can also be used to help “manage the matrix”—even areas of low current or low potential density—to ensure the widest possible

9 A large proportion of the coast has been closed to legal grizzly bear hunting since the moratorium of 2000. Proposed Benchmark Grizzly Bear Management Areas (GBCS 1995) at the Ahnuhati/Kwalate/Upper Kakweiken/Ahta and from the Dean, through the Kimsquit to Fjiordland and the Khutze have been kept closed to hunting pending the GBMA decisions. The area around Owikeno Lake has been kept closed because of conservation concerns related to the killing of problem bears at Owikeno village and the localized collapse of the Owikeno Lake sockeye run. The Grizzly Bear Management Area proposal for the area surrounding the Khutzeymateen Grizzly Bear Sanctuary, much of it closed since 1984, has been endorsed by the Kalum Land and Resource Management Plan (LRMP), tentatively endorsed by the North Coast LRMP, and is now pending government to government discussion with First Nations.



Page 67

natural distribution and persistence of the species. Refer to Appendix 1.1.2.3.5 for a summary of model peer review comments.

2.2.8 Focal Species: Black Bear (Ursus americanus) Habitat Suitability and Capability Model


Black bears are commonly found throughout coastal British Columbia. They are very adaptable and inhabit a wide variety of habitats—ranging from coastal estuaries up to high elevation alpine meadows (RIC 1999). Beaches, open riparian areas, warm aspect avalanche tracks, slides, and clearcuts are important feeding areas. They are omnivorous and opportunistic in their feeding habits. Grasses, sedges, and horsetails form the bulk of their diet, particularly in late spring and early summer. They also feed on insects, fruits, berries, fish, garbage, carrion, and small mammals and will occasionally prey on young/small deer. In the late summer and fall, salmon spawning rivers and streams represent important feeding areas (Reimchen 1998a, 1998b). Black bear habitat use is strongly influenced by intraspecific social interactions and the presence and activities of people. Feeding and security habitats in close proximity are assumed to be the limiting factors for black bears (RIC 1999).

In coastal British Columbia, black bears are old-growth structure dependent for denning (Davis 1996). Winter dens are located in large standing live, standing dead, or in or under large dead tree boles. Den cavities are dry and secure from predators (cougars, wolves, other bears). The Kermode bear, also called the Spirit Bear or Ghost Bear, is a rare white-phase black bear that also inhabits coastal British Columbia. The white genotype occurs at varying frequencies along the Northwestern coast (Ritland et al. 2001).

Black bears are globally ranked as G5—meaning they are demonstrably widespread, abundant, and secure, and provincially Yellow-listed meaning they are apparently secure and not at risk of extinction. They are also ranked by the Committee On the Status of Endangered Wildlife in Canada (COSEWIC) as “Not At Risk” (NAR) meaning they are a species that has been evaluated and found to be not at risk.

Black bears were chosen as focal species outside of occupied grizzly bear habitat for the CIT ESA because they also can be considered keystone, indicator and umbrella species (see grizzly bear model). In addition, the Kermode genotype is extremely rare, more appropriatly examined as a focal species than a special element in the ESA. Black bears were not modelled inside occupied grizzly bear habitat because of their high overlap in habitat requirements. Site series representivity was assumed to address the concern about long-term old-growth supply for denning.

2.2.8.2 Methods

The structure of the black bear model was identical to the grizzly bear model with the following exceptions:

1. Broad ecosystems were rated from the perspective of black bear habitat opportunity;



Page 68

2. Black bears show higher tolerance to roads, therefore the suitability reductions for the various road categories were adjusted as follows: suitability under road densities of 0.6 to 1.2 km/km2 was reduced by 25% and >1.2 km/km2 by 50%.


The primary model outputs are the GIS coverages and associated databases of current minimum effectiveness classes for each of the modelled watersheds in occupied black bear range within the study area (Map 11). Refer to Appendix 1.1.2.3.6 for a summary of model peer review comments.

The coarse filter biodiversity / representivity elements of the EBM Handbook will capture the key habitat supply concern for black bears in coastal BC - the provision of den trees large enough to keep bears dry and warm through the winter. Food supply (including access to salmon) and habitat security (mortality risk reduction through access management) will be covered by managing for grizzly bears in areas of overlap.

Research has demonstrated a very high degree of niche overlap. Where the fine filter of grizzly bear conservation doesn't meet the black bear requisites, the coarse filter of managing for ecosystem representivity (including within stand retention) and focused hydroriparian conservation will.

2.3 Freshwater Targets


Salmonids aside, there was very little spatially explicit special element data available for freshwater species in the study area. In Cannings and Ptolemy (1998), only 1 of the 38 rare freshwater fish species listed occurs within the study area –the giant black stickleback (Gasterosteus species 1). Four freshwater fish species were selected as special elements based on the selection criteria identified in Table 2.1: green sturgeon (Acipenser medirostris), threespine stickleback (Gasterosteus aculeatus), giant black stickleback (Gasterosteus species 1), and bull trout (Salvelinus confluentus). We were only able to obtain spatial data for the Giant black stickleback, a G1 endemic (2 element occurrence records). Three amphibian species were selected as special elements: tailed frog (Ascaphus truei), western toad (Bufo boreas), and northern red-legged frog (Rana aurora), but only the tailed frog (Ascaphus truei) (a CIT focal species) had sufficient spatial data regarding the local of its populations.

2.3.2 Representation Analysis


Freshwater ecosystems consist of a group of strongly interacting freshwater and riparian/nearshore communities held together by shared physical habitat, environmental regimes, energy exchanges, and nutrient dynamics. Freshwater ecosystems vary in their spatial extent, have indistinct boundaries, and can be hierarchically nested within one another depending on spatial scale (e.g., headwater lakes and streams are nested within larger coastal river systems). Perhaps the most distinguishing features of freshwater ecosystems from terrestrial ecosystems are their variability in form and their dynamic nature. Freshwater ecosystems are



Page 69

extremely dynamic in that they often change where they exist (e.g., a migrating river channel) and when they exist (e.g., seasonal ponds) in a time frame that we can experience. Freshwater ecosystems are nearly always found connected to and dependent upon one another, and as such they form drainage networks that constitute even larger ecological systems. They exist in many different forms, depending upon their underlying climate, geology, vegetation, and other features of the watersheds in which they occur. In very general terms, however, freshwater ecosystems fall into three major groups: standing-water ecosystems (e.g., lakes and ponds); flowing-water ecosystems (e.g., rivers and streams); and freshwater-dependent ecosystems that interface with the terrestrial ecosystems (e.g., wetlands and riparian areas).

Freshwater ecosystems support an exceptional concentration of biodiversity. Species richness is greater relative to habitat extent in freshwater ecosystems than in either marine or terrestrial ecosystems. Freshwater ecosystems contain approximately 12% of all species, with almost 25% of all vertebrate species concentrated within these freshwater habitats (Stiassny 1996). The richness of freshwater species includes a wide variety of plants, fishes, mussels, crayfish, snails, reptiles, amphibians, insects, microorganisms, birds, and mammals that live beneath the water or spend much of their time in or on the water. Many of these species depend upon the physical, chemical, and hydrologic processes and biological interactions found within freshwater ecosystems to trigger their various life cycle stages (e.g., spawning behavior of a specific fish species might need to be triggered by adequate flooding at the right time of the year, for a sufficient duration, and within the right temperature range, etc.; seed germination of a particular plant might require a different combination of variables).

Freshwater ecosystems support almost all terrestrial animal species since these species depend on freshwater ecosystems for water, food and various aspects of their life cycles. In addition, freshwater ecosystems provide environmental services such as electricity, drinking water, waste removal, crop irrigation and landscaping, transportation, manufacturing, food source, recreation, and religion and sense of place, that form the basis of our economies and social values.

Finally, all water that runs through freshwater ecosystems ends up in the ocean. There, the endless cycle of life and death in freshwater ecosystems provide a steady stream of incoming food, sediment and nutrients to estuary and nearshore marine environments. The diversity of life in the world’s freshwaters and the sea are thus closely linked as well.

2.3.2.1.1 Classification of coarse-scale freshwater ecosystems

The classification of freshwater ecosystems is a relatively new pursuit and as such this is the first attempt at a coarse-scale freshwater ecosystem classification within BC. For classification purposes, coarse-scale freshwater systems are defined as networks of streams, lakes and wetlands that are distinct in geomorphological patterns, tied together by similar environmental processes (e.g., hydrologic and nutrient regimes, access to floodplains) and gradients (e.g., temperature, chemical and habitat volume), occur in the same part of the drainage network, and form a distinguishable drainage unit on a hydrography map. Coarse-scale freshwater systems are spatially nested within major river drainages and ecological drainage units (EDUs), and are spatially represented as watershed units (specifically BC Watershed Atlas third order watersheds). They are defined at a spatial scale that is practical for regional planning. Freshwater systems provide a means to generalize about large-scale patterns in networks of streams and lakes, and the ecological processes that link them together as opposed to fine-scale freshwater



Page 70

systems, which capture a detailed and often complex picture of physical diversity at the stream, reach and lake level.

2.3.2.2 Methods

The types and distributions of coarse-scale freshwater systems are characterized based on abiotic factors that have been shown to influence the distribution of species and the spatial extent of freshwater community types. This method aims to capture the range of variability of freshwater system types by characterizing different combinations of physical habitat and environmental regimes that potentially result in unique freshwater ecosystem and community types. It is virtually impossible to build a freshwater ecosystem classification founded on biological data given that freshwater communities have not been identified in most places, and there is generally a lack of adequate survey data for freshwater species. Given that freshwater ecosystems are themselves important targets for conservation because they provide a coarse filter target and environmental context for species and communities, a classification approach that identifies and maps the diversity and distribution of these systems is a critical tool for comprehensive conservation and resource management planning. An additional advantage of such an approach is that data on physical and geographic features (hydrography, land use and soil types, roads and dams, topographic relief, precipitation, etc.), which influence the formation and current condition of freshwater ecosystems, is widely and consistently available.

The proposed freshwater ecosystem classification framework is largely based The Nature Conservancy’s classification framework for aquatic ecosystems (Higgins et al. 2002). The framework classifies environmental features of freshwater landscapes at two spatial scales. It loosely follows the hierarchical model of Tonn (1990) and Maxwell et al. (1995). It includes ecological drainage units that take into account regional drainage (zoogeography, climatic, and physiographic) patterns, and mesoscale units (coarse-scale freshwater systems) that take into account dominant environmental and ecological processes occurring within a watershed.

Seven abiotic variables were used to delineate coarse-scale freshwater system types that capture the major abiotic drivers of freshwater systems: drainage area, underlying biogeoclimatic zone and geology, stream gradient, dominant lake/wetland features, glacial connectivity and coastal connectivity. Table 2.13 summarizes data sources and variable classes for each of the classification variables. These variables are widely accepted in the literature as being the dominant variables shaping coarse scale freshwater systems and their associated communities and also strongly co-varying with many other important physical processes (i.e., Vannote et al. 1980; Mathews 1998; Poff and Ward 1989; Poff and Alan 1995; Lyons 1989; Hart and Finelli 1999; Lewis and Magnuson 1999; Newall and Magnuson 1999; Brown et al. 2003). Changing the characteristics of any of these variables for a particular coarse-scale freshwater system type will likely result in a change in freshwater communities present. For example, a steep gradient headwater system in alpine tundra running through basalt will have a very different set of freshwater communities than the same system running through limestone.



Page 71

Table 2.13 Summary of data used in coarse-scale freshwater ecosystem classification

Variable Data source(s) Variable classes

Drainage area BC Watershed Atlas, 1:50,000

< 100 km2 (headwaters, small coastal rivers)

100–1,000 km2 (small rivers)

1,000–10,000 km2 (medium rivers)

> 10,000 km2 (large rivers)

Biogeoclimatic zones

BC Ministry of Forests (2002)

Coastal Douglas-fir Zone

Coastal Western Hemlock Zone

Mountain Hemlock Zone

Bunchgrass Zone

Ponderosa Pine Zone

Interior Douglas-fir Zone

Interior Cedar–Hemlock Zone

Montane Spruce Zone

Sub-Boreal Pine–Spruce Zone

Sub-Boreal Spruce Zone

Engelmann Spruce–Subalpine Fir Zone

Boreal White and Black Spruce Zone

Spruce–Willow–Birch Zone

Alpine Tundra Zone

Bedrock geology Geology sub-classes were delineated based on the following characteristics: sediment texture; degree of weatherability/erodability; stream substrate material; and aquifer potential.

BC Ministry of Energy and Mines at 1:250,000

Bedrock geology class - subclass

Sediments - Undivided

• Chemical sediments

• Fine clastics (shale, mudstone)

• Sandstones

• Coarse clastics

• Carbonates

• Interbedded limestone/shale

Volcanics - Undivided

• Intermediate to felsic/bimodal

• Mafic

• Mixed sediments and volcanics

Intrusives - Undivided

• Intermediate to felsic

• Mafic/ultramafic

• Alkalic

Metamorphics - Undivided

• Alluvium - Till

Stream gradient BC Watershed Atlas, 1:50,000 & BC 25 m DEM

Shallow (>=50% of mainstem reaches with gradient <0.005)

Moderate (>=50% of mainstem reaches within 0.005–0.20)

Steep (>=50% of mainstem reaches with gradient >0.20)



Page 72

Variable Data source(s) Variable classes

Lakes and wetlands

BC Watershed Atlas, 1:50,000

Presence of dominant lake and wetland features if >=5% of drainage area encompassed by lake/wetland features.

Coastal connectivity


Coastally connected

Not coastally connected

Glacially fed rivers


Glacially connected (freshwater system is touching a glacier)

Glacially influenced (greater than 1% of upstream headwaters consist of glacier or permanent snow)

Not glacially connected

The freshwater classification was stratified by ecological drainage units to capture broad scale freshwater zoogeographic, physiographic and climatic patterns within each ecological drainage unit (EDU). Within each EDU, headwater, small coastal river, small river, intermediate river, and large river watersheds were identified. Watersheds within each of these drainage classes were further classified according to the dominant biogeoclimatic zone they fell within, their dominant underlying geology and their dominant stream gradient class. Each of these coarse scale freshwater system types were then further subdivided based on their characteristics of being glacially and/or coastally connected, and if dominant lake and wetland features were present.


Skeena, Nass, Haida Gwaii, and North and Central Coast EDUs collectively consist of 4,476 coarse-scale freshwater ecosystems. Table 2.14 summarizes the classification of these freshwater ecosystems into system types within each of the EDUs. Map 12 spatially summarizes the abundance and distribution of these coarse-scale freshwater system types within each of the EDUs.

Table 2.14. Summary of coarse-scale freshwater system types by EDU.

Nass Skeena North Coast Central Coast Haida Gwaii

Total number of freshwater ecosystems

755 1,333 1,773 496 125

Total number of freshwater system types

74 182 150 68 30

System types by drainage class:

Small coastal rivers

0 0 34 15 13

Headwaters 43 109 71 33 0

Small rivers 26 62 36 17 9

Intermediate rivers

4 10 9 3 8

Large rivers 1 1 0 0 0



Page 73

The resultant classification framework yields a standard reporting unit of coarse-scale freshwater system types that can be replicated across ecological drainage units. Standardizing the way coarse-scale freshwater system types are defined enables both abundance and distribution queries to be made for each freshwater system type across the Province or for a specified region. Questions related to system rarity, endemism, level of imperilment and range extent can therefore be addressed. This is helpful from a conservation planning perspective in that goals can be set differentially for different system types.

Freshwater ecosystem types derived from this assessment have value beyond supporting priority setting for biodiversity conservation. Freshwater ecosystem types can be used for evaluating and monitoring ecological potential and condition, predicting impacts from disturbance, and defining desirable future conditions. In addition, they can be used to inform sampling programs for biodiversity assessment and water quality monitoring, which requires an ecological framework in addition to a spatial framework to stratify sampling locations (Higgins et al. 2003).

We realize that this classification framework is a series of hypotheses that need to be tested and refined through additional data and expert review. We recommend that concurrently, data be gathered to refine/test the classification to bring the scientific rigor needed to further its development and use by conservation partners and agencies.

2.3.3 Focal Species: Tailed Frog (Ascaphus truei) Habitat Suitability Model


The tailed frog (Ascaphus truei), is a highly localized, specialized species that lives in cool, swift, permanently flowing headwater mountain streams composed of cobble and anchored boulders that provide refuge for tadpoles and adults. The tailed frog is exclusively a North American species. In BC—the only Canadian province where it occurs—the tailed frog is found along the Coast Mountains, from the Lower Mainland to Portland Canal, north of Prince Rupert. Although its range in BC is extensive, there are concerns about the status of the tailed frog due to its low reproductive rate, its highly specialized habitat requirements, the nature of human activities within its range and our lack of knowledge about minimum viable population size, particularly in fragmented landscapes.

The tailed frog retains more primitive characteristics than any other living frog and is the only member of the family Ascaphidae. With a life span of 15 to 20 years, the tailed frog is one of the world’s longest-lived frog species. It also has the longest larval (tadpole) phase and takes longer to reach sexual maturity than any other North American frog species. The tailed frog reaches sexual maturity at 8 or 9 years of age, which is roughly twice that of other species and results in an extremely low reproductive rate.

The tailed frog is listed as a species of special concern by the Committee on the Status of Endangered Wildlife in Canada (Dupuis 1999) and is blue listed (considered to be vulnerable) in BC. Stable headwater mountain systems with step pool channel morphology and low levels of bedload movement provide optimal habitat for this species (Dupuis and Friele 2003). These habitat characteristics are sensitive to land use practices that introduce sediment into and destabilize the stream channel (Dupuis and Steventon 1999). The species may also be negatively affected by the conversion of old growth to intensively managed second-growth forest (Corn and Bury 1989). Adult tailed frog abundance is positively correlated with the percentage of old-



Page 74

growth forest in a watershed (Dupuis and Friele 2002; Stoddard 2002; Welsh and Lind 2002), most likely because these forests dampen microclimatic extremes (Chen et al. 1993; 1995; Brosofske et al. 1997; Johnston and Frid 2002). Negative effects of forest harvesting activities on population densities and relative abundance of larvae have been documented by several researchers including Bury et al. (1991), Gilbert and Allwine (1991), Welsh and Lind (1991), and Dupuis and Steventon 1999. Corn and Bury (1989) found a lower frequency of tailed frog occurrences in logged watersheds than in unlogged ones. Due to the species’ potential sensitivity to logging it is listed as an Identified Wildlife Species in the Forest Practices Code of BC (BC Ministry of Forest and BC Ministry of the Environment 1999).

The tailed frog is at risk because it has very specialized habitat requirements, and because human activities, particular timber harvesting, are destroying its habitat. Roughly 75% of the watersheds in BC have been subject to some level of timber harvesting. Small headwater streams usually contain no fish, so they and their forested borders receive little or no protection from resource agencies and industry, unless they are unstable or have an impact on fish or their downstream spawning grounds and nurseries. The result can be extensive disturbance to tailed frog habitat, making this species vulnerable to population declines and local extirpation.

2.3.3.2 Methods

Tailed frog populations in BC have been poorly surveyed because of the remote nature of its habitat. Therefore, little is known about the viability of tailed frog population’s in BC. This model is a first of its kind in an attempt to model optimal and suitable habitat for this species at a landscape scale based on what is currently known about its local habitat. The model is based on research by Ascaphus Consulting (Dupuis and Friele 2003) and Sutherland et al. (2001) as well as direct consultation with Pierre Friele and Linda Dupuis, renowned experts on the tailed frog in BC.

Five biophysical conditions were identified as being important for tailed frog habitat:

1. Basin area between 0.3 and 10 km2. tailed frogs in this region had the highest frequency of occurrence in drainage basins within this range. These small basins represent moderate to steep tributaries with coarse substrates that feed sub-basin mainstems (Dupuis and Friele 2003).

2. Basins where the bottom elevation < 600 m and the ratio (top elevation – 900)/(900 – bottom elevation) = between 0.0 and 2.0. Dupuis and Friele (2003) found that streams with elevations less than 900 m were found to be most productive, with the highest numbers and widest distribution of cohorts on streams with low disturbance intervals. Tailed frogs in this region had the highest frequency of occurrence in the back-end of valleys. The disturbance regimes in valley heads in this region are too intense to support tailed frog populations (Dupuis and Friele 2003).

3. Watershed ruggedness between 31 and 90% where ruggedness = H/A1/2 with H = basin relief and A = basin area. Moderately steep to very steep basins mobilize sediments more readily than shallow basins providing important coarse boulder and cobble substrates for tadpole refuge and food sources. Larval densities are highest in streams dominated by boulders and cobbles, and lowest or absent in stream channels dominated by final sand and gravels (Dupuis and Friele 1996; Adams and Bury 2000; Stoddard 2002).



Page 75

4. Northerly aspect in Coast and Mountains region provides more stable flow regimes than southerly aspects which are prone to flashy flows which scour eggs and larvae (Metter 1968); Southerly aspect in Interior regions provides a more stable thermal winter regime than northerly aspect which stresses the thermal tolerance of this species (Claussen 1973; Brown 1975).

5. Forest cover age class >= 6 as an indicator of stable riparian and watershed cover. The presence of riparian buffers and old forests within a watershed are positively correlated with larval tailed frog abundance (Dupuis and Steventon 1999; Stoddard 2002; Welsh and Lind 2002). Dupuis and Friele (2002), Stoddard (2002), and Welsh and Lind (2002) have indicated that the percentage of old forest in a watershed is also positively correlated with adult abundance. It is likely that the increased structural complexity of old forests maintains a cooler microclimate for adult tailed frogs to forage in. coastal tailed frogs have been located up to 500 m from perennial streams (Dupuis and Friele 2003).

These conditions were spatially modelled within the CIT study area using the following data sets: British Columbia Watershed Atlas; Small Scale Predictive Ecosystem Mapping; Digital Elevation Model; and CIT forest cover.

An area of known western range limit on the Queen Charlotte Strait was also manually excluded. Habitat areas meeting all five biophysical conditions stated above were classified as being optimal habitat areas for tailed frog. Habitat areas meeting biophysical criteria one to three were classified as being suitable habitat areas for tailed frog.

Field count data compiled by Pierre Friele and Leo Frid; and time constrained search data from 1994 to 1998 and additional data from 2002 (Dupuis and Friele 2003) were used to verify the habitat model. These surveys were conducted in accordance with RIC Standards (BC Ministry of Environment and BC Ministry of Forests 2001).

For the purposes of assessing conservation value across the CIT study area, a conservation weight of 2 was assigned to optimal tailed frog habitat stream reaches and a conservation weight of 1 was assigned to suitable habitat stream reaches.


In total, 4,466 km of suitable tailed frog stream habitat was found to exist within the CIT study area consisting of 5,155 habitat areas. Of this suitable habitat, 2,323 km of stream habitat consisting of 2,486 habitat areas were determined to be optimal habitats that meet all five biophysical parameters within the model. Agreement between the existing field data and the model depended on how many model criteria were applied. When all criteria were in place there was a low correlation between field data and those stream reaches defined as optimal tailed frog habitat. This model may be an accurate predictor of optimal habitat, but it cannot predict probability of habitat use. It is also likely that few of the field survey sites were optimal habitat. There was a 60% correlation between survey data and suitable habitat for the tailed frog.

Linda Dupuis, Pierre Friele, Ken Dunsworth, and members of CIT’s Ecological Assessment Team reviewed the model. It was concluded by all reviewers that the model performed well in defining suitable and optimal tailed frog habitat. Map 13 represents optimal and suitable habitat defined for the tailed frog within the CIT study area.



Page 76

This habitat suitability model is a critical step in helping us to understand the scope and scale of tailed frog habitat and hence where viable populations of this species may occur within BC. This information is critical for establishing Wildlife Habitat Areas for this species.

2.3.4 Focal Species: Pacific Salmon and Steelhead


Six species of salmon occur in BC; five species of Pacific salmon—chinook (Oncorhynchus tshawytscha), chum (Oncorhynchus keta), coho (Oncorhynchus kisutch), pink (Oncorhynchus gorbuscha), and sockeye (Oncorhynchus nerka)—and steelhead (Oncorhynchus mykiss). These species have some of the most complex life histories of any species on Earth. They are wide-ranging, migratory species with life histories that integrate marine, freshwater, and terrestrial ecosystems. They carry important genetic diversity at the population level and require connectivity between life history stages and habitats across multiple biological, geographic and temporal scales. Subsequently, they face critical threats at all scales, across all life history stages and habitats. They are considered a key set of focal species for the Ecosystem Spatial Assessment not only because of their highly specialized life histories but also because they play a critical role in the integrity of BC’s coastal ecosystems.

BC’s coastal ecosystems evolved in conditions where nutrients were transferred from the marine environment to freshwater and terrestrial environments through the annual return of spawning salmon. The nitrogen, phosphorous, and carbon delivered to rivers, estuaries and riparian zones through predation and decomposition of spawned salmon form important biochemical building blocks of these ecosystems (Cederholm et al. 1989; Kline et al. 1990; Bilby et al. 1998; Ben-David 1998; Schmidt 1998; Finney et al. 2000). Helfield and Naiman (2001) concluded that this fertilization process serves not only to enhance riparian production, but may also act as a positive feedback mechanism by which salmon borne nutrients improve spawning and rearing habitat for subsequent spawning generations, and maintain the long-term productivity of river corridors along BC’s coast. Gresh et al. (2000) suggested that between 120 million and 260 million kg of salmon biomass once returned to BC’s rivers. Today, this number is estimated at 60 million kg, a 50–75% decline in salmon biomass and a nutrient deficit of between 2 and 6 million kg of nitrogen and phosphorous.

Salmon are also critical to coastal ecosystems because they are a desirable prey species. Their existence within coastal BC ecosystems plays a major role in the dynamics of wildlife and regional biodiversity (Wilson and Halupka 1995; Ben-David et al. 1998). The range of organisms using salmon carcasses extends from microscopic invertebrates to large vertebrate carnivores. The survival of these species is often linked to the timing of the salmon runs (Cederholm et al. 2000). Spawning salmon are a major food source for grizzly and black bears along the coast of BC especially during fall preparation for winter denning when salmon consist of over 90% of coastal bears’ diet (Hildebrand et al. 1996). Salmon are also the predominant prey of resident killer whales in coastal BC and adjacent waters. Over 95% of documented predation events by resident Orcas are on salmon (Ford et al. 1998).

Although each of these six species of salmon shares unique evolutionary characteristics within the genus Oncorhynchus, they each exhibit unique life history characteristics and as such should be assessed independently. Furthermore, based on the genetic distinctiveness of even and odd



Page 77

year pink salmon and summer and winter run steelhead, these four species types were assessed independently.

2.3.4.2 Methods 2.3.4.2.1 Pacific salmon

The federal Department of Fisheries and Oceans (DFO) salmon escapement database was used to assess trends in salmon escapement over time and compare biomass estimates across runs. DFO enumerates salmon escapement (number of salmon returning to their natal streams to spawn) annually, an enormous undertaking given the size and geography of coastal BC and the multiple species and river systems. Methods used to count fish include permanent fences, visual estimations, fish wheels, aerial counts and dive surveys. While the escapement database is important, it has its limitations. Knowledge of particular river systems and species range from extensive to none existent. The present database contains escapement estimates for Pacific salmon from the 1950s to the present. It would have been tremendously beneficial to have had a database that started prior to the industrial revolution given that activities such as commercial fishing, logging, and watershed development were already extensive by then. Adverse weather, changes in surveyor/stream walkers, and inconsistent methodology can all lead to misrepresentation of the real trends in escapement, and may mask the effects of other influences such as logging and land use, over-fishing, climatic and natural variations. Lastly, although theoretically only native individuals are enumerated upon their return to their natal streams, offspring that are the result of successful reproduction of hatchery and native stock are counted in these enumerations, which may artificially increase escapement values. Regardless of its problems, DFO’s salmon escapement database contains the best available survey information on the state of BC’s salmon.

Runs with less than ten years of data were removed from the analysis and labeled as having insufficient data. Trends in escapement over time were calculated for each run using log (e) +1 escapement over years. Runs were classified as exhibiting either a stable or declining escapement trend. (Note that an in-depth statistical analysis of long and short-term escapement trends for each run and verification of the data is currently underway but was not completed in time for incorporation into this study.) Runs with slopes greater than or equal to 0.0 were characterized as having stable escapement trends over time. Runs with slopes less than 0.0 were characterized as having declining escapement trends over time. Biomass estimates for each run were calculated using the median of ten year running averages for each run. Average weights for each species were determined using Gresh et al. (2000).

2.3.4.2.2 Steelhead

Steelhead populations are managed by the Province of British Columbia using a different management framework from DFO. Population trends were estimated from catch information collected annually from a mail out survey, from test fisheries in commercial fisheries with a steelhead bycatch, and from swim surveys, weir counts, and creel surveys. Population status was assigned by regional biologists using various combinations of the above methods. Results were expressed as the number of spawners divided by the estimated carrying capacity of the river. Populations were assigned to one of the following categories:

1. Routine Management Zone (RMZ) (> 30% of carrying capacity)



Page 78

2. Conservation Concern Zone (CCZ) (between 10 and 30% of carrying capacity)

3. Extreme Conservation Concern (ECZ) (less than 10% of carrying capacity.

For the Ecosystem Spatial Analysis, stable populations were considered to be populations in the RMZ, while the remaining zones were considered to be equivalent to the declining populations in the Pacific salmon analysis.

Biomass calculations for steelhead were rough estimates designed to place the populations in categories of large, medium and small populations using a log10 scale based on known or estimated population sizes. Missing values were then estimated by interpolation.

2.3.4.2.3 Salmon population units

Salmon population units (SPUs) were derived for each of the six salmon species based on existing FISS point source data, DFO escapement data, and experts knowledge of runs sharing similar genetics and habitat preferences. Escapement trends and biomass estimates were summarized for each SPU by synthesizing run data where appropriate. Biomass estimates were summed for runs within a SPU. A SPU was characterized as containing high biomass runs if accumulative runs were >= 2.0 on a log (10) scale and low biomass if they were less than 2.0 on a log (10) scale. Trends in escapement over time were examined across all runs within a SPU. If a stable or declining trend across runs was not apparent, then biomass was examined in conjunction with slope trend and the slope trend for the run(s) with the highest biomass were selected as being representative of the SPU. It should be noted that for the majority of the SPUs, trends in escapement over time were very obvious and biomass estimates did not need to be factored into the determination of slope trends.

2.3.4.2.4 Goal setting

From a biodiversity standpoint all salmon SPUs are important and equally valued. However, for the purpose of priority setting for conservation, the following weightings were used to set conservation goals: stable SPUs with high biomass were given three times the conservation weight than declining SPUs with low biomass as well as stocks with insufficient data. Stable SPUs with low biomass and declining SPUs with high biomass were given a conservation weighting of two (Table 2.15). These conservation goals for each salmon species were stratified by ecological drainage units (EDUs) and used in SITES algorithm as input files for determining areas of high conservation value across the CIT study area.



Page 79

Table 2.15 Summary of weightings used to set conservation goals for all salmon SPUs


All Pacific salmon species showed a serious decline in escapement over time with 40% of these declines occurring in high biomass runs. Pink (even year) runs were found to be in the best condition and Chum in the worst condition. Steelhead runs were found to be in stable condition across all EDUs except for Central Coast. Of all EDUs within the CIT region, Central Coast was found to in the worst condition with 80% of its salmon SPUs in decline. Haida Gwaii was a close second with 76% of its SPUs in decline. Maps 14 to 21 summarize the spatial trends in escapement and biomass for each of the six salmon species across the CIT study area. For a detailed summary of escapement and biomass trends by EDU, please refer to Tables 2.16 to 2.21. For a detailed summary of combined escapement and biomass trends for each species, refer to Figures 2.1 to 2.9. The following sections highlight general findings of trends across EDUs as well as species trends.

2.3.4.3.1 Trends across Ecological Drainage Units (EDUs)

Salmon in the Skeena River EDU were generally in the best condition of all EDUs within the study area with 40% of its Pacific salmon SPUs containing stable runs and 100% of its steelhead SPUs containing stable runs. The majority (69%) of all SPUs in the Skeena EDU were of low biomass except for sockeye, which had, 64% of its SPUs composed of high biomass runs.

Sixty-three percent of Pacific salmon SPUs within the Nass EDU were in decline. Coho appeared most stable with 53% of its SPUs containing stable runs. All summer and winter steelhead SPUs were stable. Majority of all species SPUs were of low biomass (58%) except for sockeye in which 83% of its SPUs were high biomass runs.

The majority (67%) of Pacific salmon SPUs were in decline within North Coast EDU. Nine-seven percent of summer steelhead SPUs were in decline compared with only 7% of steelhead winter SPUs. The majority of Pink (even and odd year) and Chum SPUs had high biomass runs.

The majority (76%) of Pacific salmon SPUs were in decline within the Haida Gwaii EDU. There is only one chinook population within Haida Gwaii that is located within the Yakoun River. It had a stable population with a high biomass run. All summer and winter steelhead SPUs were stable. The majority (52%) of all species’ SPUs had high biomass runs except for Pink (odd) and



Page 80

steelhead (summer and winter) with 85% and 100%, respectively, of their SPUs scoring as low biomass.

The Central Coast EDU was found to be in the worst condition of all EDUs with 80% of its Pacific salmon SPUs and 48% of its steelhead (summer and winter) SPUs in decline. The exception was pink (even year) which had 71% of its SPUs in stable condition within this EDU. The majority (66%) of all species’ SPUs had low biomass runs except for Pinks (even year) with 54% of their SPUs scoring as high biomass runs.

2.3.4.3.2 Species trends

All salmon species were found to have greater than two-thirds of their SPUs in decline except for even year Pink (53% in decline) and steelhead (33% winter and 12% summer in decline). Chum had the highest proportion of its SPUs in decline (77%) followed closely by coho, sockeye, chinook, and pink (odd). Pink (even year) had the highest proportion of its SPUs composed of high biomass runs (72%) followed by sockeye (55%). The remaining species had the majority of their SPUs composed of low biomass runs. The following is a summary of each salmon species by EDU.

Sockeye SPUs were found to be in the best condition in Skeena (54% of its SPUs with stable escapement trends) and Nass (84% of its SPUs with stable escapement trends). However, 74% of all sockeye SPUs within the CIT region were found to be in decline with the majority of these being high biomass runs.

Pink (even year) were found to be in the best condition of all Pacific salmon species with 47% of its SPUs in stable condition. Central Coast (71%), and Skeena (54%) EDUs were found to contain the highest proportion of pink (even year) SPUs in stable condition. For the 53% of pink (even year) SPUs that were in decline within the CIT region, the majority of them were composed of high biomass runs.

Pink (odd year) were found to be in the second best condition of all Pacific salmon species with 31% of its SPUs in stable condition. Central Coast (92%) and Haida Gwaii (96%) were found to contain the highest proportion of pink (odd year) SPUs in stable condition. For the 69% of the Pink (odd year) runs that were found to be in decline within the CIT region, the majority of them were low biomass runs.

Sixty-nine percent of all chinook SPUs and 76% of all coho SPUs were found to be in decline within the CIT region. Of these chinook and coho SPUs, the majority were composed of low biomass runs. Chinook expressed a decline in escapement trends across all EDUs except for Skeena where just over half of its SPUs (52%) were found to be in stable condition. Coho expressed a decline in escapement trends across all EDUs except for Nass where just over half of its SPUs (53%) were found to be in stable condition.

Seventy-seven percent of all chum SPUs were found to be in decline within the CIT region. Of these chum SPUs, the majority were composed of low biomass runs. This decline in escapement trend was prominent across all EDUs. These results indicate that chum may be in the worst shape of all six salmon species analyzed in this study.



Page 81

Summer and winter run steelhead were found to be in the best condition of all six salmon species. One hundred percent of their SPUs were stable in the Nass, Skeena, and Haida Gwaii EDUs. They were in greatest decline in the Central Coast EDU with 100% of summer run SPUs and 45% of winter run SPUs in decline. Summer and winter run steelhead had consistently low biomass runs across all EDUs except for summer run steelhead in the Nass (31%) and Skeena (42%) EDUs.

Table 2.16 CIT regional summary of salmon status

Analysis Pink (Even)

Pink (Odd) Chum Chinook Coho Sockeye Winter

steelhead Summer

steelhead

Total Salmon Conservation Units

384 384 309 139 488 217 664 112

Salmon Conservation Units with Sufficient Data

232

(60%)

356

(93%)

216

(70%)

81

(58%)

260

(53%)

99

(46%)

664

(100%)

112

(100%)

Trends Based on Sufficient Data

% Stable 47 31 23 31 24 28 67 88 Escape-ment Trends %

Declining 53 69 77 69 76 74 33 12

% High Biomass

72 43 48 24 30 55 19 2 Biomass Trends

% Low Biomass

28 57 52 76 70 45 81 98

Table 2.17 Summary of salmon status for Nass River EDU



steelhead Summer

steelhead

Total Salmon Conservation Units with Sufficient Data

10 21 6 11 17 6 30 13

% Stable 40 38 0 27 53 33 100 100 Escapement Trends

% Declining

60 62 100 73 47 67 0 0

% High Biomass

50 43 17 36 35 83 10 31 Biomass Trends

% Low Biomass

50 57 83 64 65 17 90 69



Page 82

Table 2.18 Summary of salmon status for Skeena River EU



steelhead Summer

steelhead


28 74 17 31 58 28 48 33


% Declining

46 73 65 18 60 46 0 0

% High Biomass

36 35 12 19 21 64 7 42 Biomass Trends

% Low Biomass

64 65 88 81 79 36 93 58

Table 2.19 Summary of salmon status for North Coast EDU



steelhead Winter

steelhead


127 146 120 24 103 39 452 34


% Declining

52 49 77 79 87 85 7 97

% High Biomass

81 64 56 21 34 46 2 9 Biomass Trends

% Low Biomass

19 36 44 79 66 54 98 91

Table 2.20 Summary of salmon status for Central Coast EDU



steelhead Summer

steelhead


24 49 38 10 38 10 95 4


% Declining

29 92 87 100 84 90 45 100

% High Biomass

54 26 45 20 26 30 2 0 Biomass Trends

% Low Biomass

46 74 55 80 74 70 98 100



Page 83

Table 2.21 Summary of salmon status for Haida Gwaii EDU



steelhead Summer

steelhead


31 45 23 1 27 12 39 28


% Declining

64 96 65 0 67 83 0 0

% High Biomass

90 16 65 100 48 67 0 0 Biomass Trends

% Low Biomass

10 84 35 0 52 33 100 100

2.3.4.4 Conclusion

Even though there is a large amount of uncertainty surrounding DFO’s salmon escapement database, the results of this study indicate robust downward trends in escapement over time for all five species of Pacific salmon throughout the entire CIT study area as well as prominent declines in winter and summer steelhead within the Central Coast. This is a serious cause for concern regarding the long-term health of BC’s salmon. Causes of decline in salmon populations over time stem from multiple factors including ocean regime patterns, marine survival conditions, climate change, changes in land use, hatcheries, and overfishing. To keep these salmon populations from extinction, we will need to implement more conservative harvest and hatchery’s management and habitat protection strategies. Land use planning tables need to consider the long-term protection of salmon habitat as a high priority if we have any hope of saving these salmon populations from extinction.



Page 84

Figure 2.1 Summary of salmon species condition within CIT study area.

Num

ber o

f sal

mon

pop

ulat

ion

units

(576)

0

5

10

15

20P

ink

(Eve

n Y

ear)

Pin

k (O

dd Y

ear)

Soc

keye

Chi

nook

Coh

o

Chu

m

Sum

mer

Ste

elhe

ad

Win

ter S

teel

head

Stable/High Biomass

Stable/Low Biomass

Declining/High Biomass

Declining/Low Biomass



Page 85

Figure 2.2 Summary of coho condition by EDU.

0 10 20 30 40 50 60 70

Nass Skeena North Coast

CentralCoast

Haida Gwaii

Stable/High Biomass Stable/Low Biomass Declining/High Biomass Declining/Low Biomass

Num

ber o

f sal

mon

pop

ulat

ion

units



Page 86

Figure 2.3. Summary of chum condition by EDU.

0 10 20 30 40 50 60

Nass Skeena North Coast

CentralCoast

Haida Gwaii


Num

ber o

f sal

mon

pop

ulat

ion

units



Page 87

Figure 2.4. Summary of sockeye condition by EDU.

0 2 4 6 8

10 12 14 16 18

Nass Skeena North Coast Central

CoastHaida Gwaii


Num

ber o

f sal

mon

pop

ulat

ion

units



Page 88

Figure 2.5. Summary of odd year pink condition by EDU.

0 10 20 30 40 50 60 70


CoastHaida Gwaii


Num

ber o

f sal

mon

pop

ulat

ion

units



Page 89

Figure 2.6. Summary of even year pink condition by EDU.

0 10 20 30 40 50 60


CoastHaida Gwaii


Num

ber o

f sal

mon

pop

ulat

ion

units



Page 90

Figure 2.7. Summary of chinook condition by EDU.

0 2 4 6 8

10 12 14 16 18


CoastHaida Gwaii


Num

ber o

f sal

mon

pop

ulat

ion

units



Page 91

Figure 2.8. Summary of summer steelhead condition by EDU.

0 5

10 15 20 25 30 35


CoastHaida Gwaii


Num

ber o

f sal

mon

pop

ulat

ion

units



Page 92

Figure 2.9. Summary of winter steelhead condition by EDU.

0

5

10

15

20

25

30

35

40

45


CoastHaida Gwaii

Stable/High BiomassStable/Low BiomassDeclining/High BiomassDeclining/Low Biomass

Num

ber o

f sal

mon

pop

ulat

ion

units



Page 93

2.4 Marine Targets

2.4.1 Nearshore Marine


This section provides assessment information about the nearshore marine ecological systems and species in Queen Charlotte Island, Gwaii Haanas, and the North and Central Coasts of British Columbia. The objective of the nearshore marine component was to identify a set of conservation areas (i.e., an ecoregional portfolio) that, if conserved, will protect a representative subset of the nearshore marine biodiversity of those waters. A technical workshop set the stage for a nearshore assessment to identify specific conservation targets in the form of marine species, shoreline types, and nearshore marine ecosystems. The workshop also identified key areas in the coastal zone known for their diverse species assemblage or high abundance of certain species. Refer to Appendix 2.1 for full details of the workshop. We set conservation goals for each marine species and ecosystem target, and ran the site selection algorithm MARXAN to compare these known significant areas and add spatially explicit information to them. The next step of this process is to conduct a thorough review of the analytical output and compare it to key areas identified in the workshop.

The Nature Conservancy’s ecoregional planning process identifies and plans for the protection of species and ecosystems. Their identification in this process leads to a thorough data collection effort, which then becomes the core ingredients for analysis. It is therefore necessary to clearly define the species and ecosystems in the marine environment that are being targeted, and the data that exist to back up these definitions.

2.4.1.2 Methods

We identified the intertidal and nearshore zone as the part of the marine environment to target and plan for protection. This zone roughly spans the supratidal area above the ordinary or mean high water line (i.e., the top of a bluff or the extent of a saltmarsh in the upper intertidal) to the subtidal area. The subtidal begins at approximately the mean lower low water line (zero feet elevation) down to the –20 m isobath. The limit on the lower extent was defined by the shore-zone mapping system (see Berry et al. 2001). This was the definition of the nearshore zone that was adopted for the regional analysis. Therefore when the following text refers to the “shoreline,” it specifically means the analysis conducted using the shorezone data, and when the text refers to “nearshore,” it means the entire nearshore zone including intertidal vegetation and habitats, forage fish spawning sites, and seabird colonies.

We identified 72 conservation targets of which 59 were coarse filter targets (shoreline ecosystems) and 13 were fine filter targets (specific occurrences of marine species). This set of targets was selected to represent the full nearshore marine biodiversity of the region, and to highlight elements that are threatened or declining (i.e., seabird species), or that serve as good indicators of the health of the larger ecosystem (i.e., intertidal habitats).

The waters of the ecoregion were divided, or stratified, into 8 sections based on British Columbia’s marine “ecosections” (Harper 1993). Ecosections are characterized as unique physiographic, oceanographic, and biological assemblages that are related to water depth and habitat (pelagic versus benthic). There are several implicit ecological factors of significance to



Page 94

ecosections, the most detailed level of classification after “ecozones,” “ecoprovinces,” “ecoregions,” and “ecodistricts.” Ecosections are an indicator of major community differences between benthic and pelagic biota (i.e., solid versus fluid). They are an indirect indicator of primary productivity where shallow areas in the euphotic zone have higher productivity. They are also a direct indicator of oceanic stratification and associated biological assemblages. We used this fifth order subdivision to stratify the coastal zone (Map 22).

In addition to dividing the coastal zone by marine ecosection, we also used a stratification scheme based on project regions. The Ministry of Sustainable Resource Management (MSRM) has contracted with several consulting firms to inventory the shoreline using the shore-zone mapping system (see next section, “Coarse Filter Targets”). Six separate project regions (Map 24) define the CIT region, with varying methodologies adopted for inventorying and interpreting the flight surveys from one region to another. As such, we attempted to account for this discrepancy by ensuring coastal representation across these regions in addition to the more ecological stratification. A third scenario was to not stratify the region at all, but run the site selection algorithm on the 72 conservation targets across the entire CIT region.

2.4.1.3 Special Elements

In selecting these targets, we had to decipher what species the coarse filter targets were not adequately representing. In general, we included species that are considered imperiled, keystone (Power et al. 1996), or an ecosystem engineer that has a major impact on the structure and function of communities. Consideration was given to whether the species met the criteria across the study region, or only within a portion of it. The latter case often occurred where a species was imperiled or listed within a certain jurisdiction, and yet the population was thriving in another section of the CIT region. When this occurred it was not appropriate to include the species in the regional analysis.

With these criteria we included 8 marine species as targets. These forage fish and seabird targets are represented as 2 critical life stages, spawning aggregations and breeding colonies.

2.4.1.3.1 Forage fish (spawning sites)

Pacific herring spawning sites were collected from MSRM. We had comprehensive coverage for herring in the study area, assembled as linear features that coincide spatially with the shorezone mapping system. Attribute data indicate the Relative Importance (RI) of the feature per location. The values are only comparable within project regions (i.e., Queen Charlotte Islands); this is currently being changed so that these values will be comparable across British Columbia. Nonetheless, we selected values between 3 and 5 (medium to very high) for the analysis. This information was attributed to the shoreline segments.

2.4.1.3.2 Seabird colonies

The Canadian Wildlife Service, who compiled the data for distribution in 2001, provided seabird colony information. Most of the data are the results of a comprehensive inventory of colonial nesting seabirds along the British Columbia coastline conducted between 1980 and 1989 by the Canadian Wildlife Service of Environment Canada. This data set includes the locations of all



Page 95

known seabird colonies along the coast of British Columbia, and provides a compilation of the most recent (up to 1989) population estimates of seabirds breeding at those colonies (Map 25).

Fifteen species of seabirds, (including two storm petrels, three cormorants, one gull and nine alcids) and one shorebird (Black Oystercatcher) breed on the coast of British Columbia. We selected those species that are considered imperiled or vulnerable in British Columbia (S1–S3 status). The source of this information was from NatureServe.10 There were 7 species of seabirds selected to represent colonies along the coast of British Columbia. These include Thick-billed and Common Murre, Cassin’s Auklet, Ancient Murrelet, Horned and Tufted Puffin, and Brandt’s Cormorant. These data are represented as point locations with attributes describing their location. We used location descriptions to assign the colonies to specific shoreline segments.

2.4.1.4 Representation

The coarse filter targets comprise shoreline ecosystem types derived from a single classification developed in British Columbia, the shorezone mapping system. The Province of British Columbia developed its physical and biological shorezone mapping system based on shore types after Howes et al. (1994) and Searing and Frith (1995). Shore types are biophysical types that describe the substrate, exposure, and vegetation across the tidal elevation, as well as the anthropogenic features. The B.C. shorezone mapping system is built on shore types that aggregate precise community or habitat types according to their landform, substrate, and slope (Berry et al. 2001). There are 34 coastal classes and 17 representative types within the classification system (Table 2.22). See Berry et al. (2001) for the rationale and definitions of the 34 coastal classes. The definitions of the 17 representative types are listed in Appendix 2.2.

We examined the 17 representative types within the classification and added the “bio-exposure” field in the data sets to form our shoreline ecosystem conservation targets. Bio-exposure is defined as a summary classification of indicator species and intertidal vegetation observed in each shoreline segment (Morris et al. 2001). Species present in turn indicated the wave exposure energy for that segment. Biologists surveying the coastline had the highest level of confidence in assigning wave energy to places where intertidal vegetation was visible. Not all project regions contained the bio-exposure field, however, but instead contained an observation of exposure interpreted from a geologist. Given substrate and exposure information one can then assume the biological assemblages. This exposure observation was deemed of medium confidence in the shorezone system, used only when bio-exposure was not available. There were 6 wave energy classes assigned to shoreline segments, ranging from very protected to very exposed. There was an additional category that did not identify wave energy.

10 NatureServe Explorer: An online encyclopedia of life [web application]. 2002. Version 1.6. Arlington, Virginia, USA: NatureServe. Available: http://www.natureserve.org/explorer



Page 96

Table 2.22 BC Coastal classes and representative types

Coastal class Description Representative type

0 Undefined Undefined

1 Rock ramp, wide Rock Platform

2 Rock platform, wide Rock Platform

3 Rock cliff, narrow Rock Cliff

4 Rock ramp, narrow Rock Cliff

5 Rock platform, narrow Rock Platform

6 Ramp with gravel beach, wide Rock with Gravel Beach

7 Platform with gravel beach, wide Rock with Gravel Beach

8 Cliff with gravel beach Rock with Gravel Beach

9 Ramp with gravel beach, narrow Rock with Gravel Beach

10 Platform with gravel beach, narrow Rock with Gravel Beach

11 Ramp with gravel and sand beach, wide Rock with Sand & Gravel Beach

12 Platform with gravel and sand beach, wide Rock with Sand & Gravel Beach

13 Cliff with gravel and sand beach Rock with Sand & Gravel Beach

14 Ramp with gravel and sand beach, narrow Rock with Sand & Gravel Beach

15 Platform with gravel and sand beach, narrow Rock with Sand & Gravel Beach

16 Ramp with sand beach, wide Rock with Sand Beach

17 Platform with sand beach, wide Rock with Sand Beach

18 Cliff with sand beach Rock with Sand Beach

19 Ramp with sand beach, narrow Rock with Sand Beach

20 Platform with sand beach, narrow Rock with Sand Beach

21 Gravel flat, wide Gravel Flats

22 Gravel beach, narrow Gravel Beach

23 Gravel flat or fan Gravel Beach

24 Sand and gravel flat or fan Sand & Gravel Flat

25 Sand and gravel beach, narrow Sand & Gravel Beach

26 Sand and gravel flat or fan, narrow Sand & Gravel Flat

27 Sand beach Sand Beach

28 Sand flat Sand Flat

29 Mud flat Mud Flat

30 Sand beach, narrow Sand Beach

31 Organics/fines Estuary Wetland

32 Man-made, permeable Man-made

33 Man-made, impermeable Man-made

34 Channel Channel



Page 97

2.4.1.4.1 Shoreline ecosystem targets

The 17 representative types and 7 wave energy classes (very protected, protected, semi-protected, semi-exposed, exposed, very exposed, plus an unidentified category) yielded 119 potential shoreline targets. We decided that having too many classes implied a level of detail that was unnecessary at the scale of the CIT region. Therefore the wave energy classes were aggregated down to 5 classes, where semi-protected was combined with the protected class, and semi-exposed was combined with exposed. We felt that this generalized the shoreline ecosystems into discernible coastal communities for planning purposes at this regional scale. This yielded as many as 85 classes depending on which wave energy classes were possible for each representative shoreline type.

In actuality there are only 63 shoreline categories available after compiling all 6 project regions of shorezone (Table 2.23). There are 62,441 distinct shoreline segments that cover the entire CIT region. The total shoreline length is 28,145 km compiled from data sets at scales ranging from 1:40,000 to 1:50,000. Four categories, 3 “man-made” types and 1 “unidentified,” are not considered ecosystem targets, bring the total number of targets down to 59. Man-made types consist of 320 shoreline segments (120 km), and unidentified shorelines consist of 1,026 segments (663 km). Removing these categories left 61,094 shoreline segments with targets, or 27,340 km.

These coarse filter shoreline types range from 8 to 21,132 m in length, with a mean of 450 m. Short shoreline segments tend to be tiny islands either within estuaries or off the coast. The minimum mapping unit (the smallest measurement that can be delineated on a map as a shape with boundaries) of the actual shoreline classification segments is estimated at 25 m when mapping is conducted using 1:50,000 hardcopy maps. There are 42 individual shoreline segments (828 m) under 25 m in the study area. Shorter segments therefore require further evaluation as to their positional accuracy. It should be noted here that the CIT was given draft shorezone data from MSRM. Although we were given access to these data to complete the coastal/nearshore analysis, there have been changes to the data sets since we acquired it. Therefore the shoreline classes and statistics associated with the data have changed since the completion of this analysis and report.

Table 2.23 Shoreline ecosystem targets

Target Target_ID Shore units Total length (m)

Channel 1001 3 1,548.5

ChannelE 1002 9 2,690.1

ChannelP 1003 73 38,751.9

Estuary Wetland 1004 40 43,949.0

Estuary WetlandE 1005 13 9,942.3

Estuary WetlandP 1006 1134 1,046,811.2

Estuary WetlandVE 1007 9 9,323.4

Estuary WetlandVP 1008 47 46,032.7

Gravel Beach 1009 215 78,725.4

Gravel BeachE 1010 333 93,229.1

Gravel BeachP 1011 2983 941,273.4



Page 98


Gravel BeachVP 1012 12 4,336.2

Gravel Flats 1013 28 15,052.7

Gravel FlatsE 1014 56 19,683.4

Gravel FlatsP 1015 148 76,247.2

Gravel FlatsVE 1016 9 3,599.3

High Tide Lagoon 1017 50 85,740.2

High Tide LagoonE 1018 10 4,254.7

High Tide LagoonP 1019 210 236,286.4

High Tide LagoonVP 1020 60 80,393.8

Man-made (not a target) 1021 13 5,253.3

Man-madeE (not a target) 1022 5 1,262.8

Man-madeP (not a target) 1023 302 135,230.8

Mud Flat 1024 7 4,985.3

Mud FlatP 1025 156 107,010.0

Rock Cliff 1026 1396 703,588.4

Rock CliffE 1027 4314 1,768,132.6

Rock CliffP 1028 11822 5,833,450.0

Rock CliffVE 1029 218 125,300.5

Rock CliffVP 1030 68 31,122.4

Rock Platform 1031 124 33,573.1

Rock PlatformE 1032 1774 661,917.3

Rock PlatformP 1033 447 128,569.6

Rock PlatformVE 1034 345 162,097.7

Rock with Gravel Beach 1035 710 287,587.8

Rock with Gravel BeachE 1036 3374 1,394,014.4

Rock with Gravel BeachP 1037 9679 4,406,751.8

Rock with Gravel BeachVE 1038 123 40,373.1

Rock with Gravel BeachVP 1039 74 35,028.4

Rock with Sand & Gravel Beach 1040 173 57,913.7

Rock with Sand & Gravel BeachE 1041 2333 876,653.2

Rock with Sand & Gravel BeachP 1042 7600 2,843,714.0

Rock with Sand & Gravel BeachVE 1043 12 11,238.0

Rock with Sand & Gravel BeachVP 1044 39 15,471.3

Rock with Sand Beach 1045 19 4,346.8

Rock with Sand BeachE 1046 363 124,830.9

Rock with Sand BeachP 1047 544 182,404.5

Sand & Gravel Beach 1048 173 59,406.9



Page 99


Sand & Gravel BeachE 1049 198 53,307.5

Sand & Gravel BeachP 1050 3600 1,339,292.9

Sand & Gravel BeachVP 1051 29 8,876.9

Sand & Gravel Flat 1052 96 38,497.7

Sand & Gravel FlatE 1053 176 99,267.0

Sand & Gravel FlatP 1054 2689 1,298,001.0

Sand & Gravel FlatVP 1055 29 12,829.3

Sand Beach 1056 19 10,393.3

Sand BeachE 1057 242 187,660.8

Sand BeachP 1058 298 124,058.7

Sand Flat 1059 42 23,653.5

Sand FlatE 1060 129 78,238.5

Sand FlatP 1061 2179 1,306,395.7

Sand FlatVP 1062 40 22,440.9

Undefined (not a target) 9999 1026 662,992.3

59 Targets 61,095 27,340,266.0

Total shoreline contains 62,441 units with a total length of 28,145,005.1 m

E = EXPOSED

P = PROTECTED

VE = VERY EXPOSED

VP = VERY PROTECTED

Note: No exposure class indicates that the exposure was not defined

2.4.1.4.2 Intertidal vegetation and habitat targets

The shorezone mapping system also identified intertidal vegetation and habitats. “Bio-bands” are assemblages of intertidal biota, visible from the air and named for dominant species or species assemblages. With these bio-bands we identified the intertidal range biota in the nearshore, including saltmarsh, eelgrass, surfgrass, and kelp. Saltmarsh consists of grasses, herbs, and Salicornia associated with protected estuaries in the high intertidal zone. Eelgrass consists of Zostera species found in protected environments of the shallow subtidal. Surfgrass, or Phyllospadix, is another seagrass found in more exposed coastlines also in the shallow subtidal. Kelps consist of both Macrocystis and Nereocystis species, where Macrocystis is found in protected areas and Nereocystis is found on more exposed rocky shores. These kelps are found in the subtidal zone. We found that these vegetation types form the major habitats of the nearshore zone, and although these types alone do not represent that most diverse intertidal habitats, they are extremely biologically productive and the most sensitive to human alteration. Further, these categories are recognized ecologically, are protected by policy, and are the best surrogates at this scale to represent a range of habitats.

We also used the “habitat observed” category from shorezone to capture the most diverse part of the intertidal zone. Habitat observed is the summary classification of the segment’s bio-exposure



Page 100

—from indicator species and bio-bands—in combination with the classification of the segment’s substrate and morphology. There are 10 different habitat categories in the data set. We used the Habitat 3 category as additional target to represent the semi-exposed, immobile bedrocks and blocks (coastal classes 1 – 23; see 4.4.1.4a) in the lower intertidal. Indictor species include rich red algae, Hedophyllum, Egregia, Phyllospadix, L. setchellii, and Eisenia. Indicator bio-bands include mussels-barnacles, chocolate brown algae, surfgrass, urchin, and Nereocystis kelp. This category represents the highest diversity of biota in the intertidal (Mary Morris, pers. comm., 2003).

2.4.1.5 Data Gaps

In conducting the CIT nearshore analysis, there are bound to be data limitations and inconsistencies across jurisdictional boundaries. In general, marine data are less developed than terrestrial counterparts. This is especially true for presence/absence data on marine algae and sessile invertebrates that define ecosystems in much the way plants do on land.

2.4.1.5.1 Fine filter data gaps

Marine species data are either very coarse in scale (i.e., depicting a species’ distribution over a large area) or collected on a very fine scale of resolution (i.e., detailed survey transects a specific site, but the coverage across a wide area is highly discontinuous and methods vary tremendously). Including this type of information in explicit site selection analysis is problematic; therefore we screened data for inclusion in the regional analysis. We assessed all available data sets to include in the regional analysis, and set different parameters during the site selection analysis (see Section 7.2.4)

Our biggest data gap in B.C. was data regarding invertebrates. The Conservation Data Centre (CDC) does collect element occurrence data for some invertebrate species, but the data is not comprehensive throughout the study area. Without a comprehensive, continuous survey effort we were limited by the places where species were found and therefore did not have a sense of abundance across the region. Although this is a systemic problem for all spatial analyses, it is particularly problematic for sessile invertebrates that may use large areas of benthic habitat types. The sparse data reflects neither the best nor the only sites where these species occur.

In B.C. we had no fishery-independent survey information on rockfish species, lingcod, surf smelt, or sand lance. Only Pacific Herring spawning data allowed for a comprehensive analysis across the region.

We had good representative data for seabird colonies, though there is some debate over what constitute as a seabird target. Depending on the source being used, different planning teams have selected different species to include in the analysis (i.e., nearshore and offshore analyses). Without a definitive study or source for selecting seabirds as conservation targets, there will continue to be debate over what species meet conservation target criteria. Although not a data “gap,” this inconsistency in the selection process have made the integration of nearshore and offshore analyses more difficult (where the nearshore has selected specific seabirds as colony habitats, the offshore analysis focused on a different list of target species and their foraging areas.)



Page 101

2.4.1.5.2 Coarse filter data gaps

Although we included “estuaries” as targets in the analysis, we did not have complete information on all estuaries in B.C. Further, we were limited by the treatment of estuaries as they were defined spatially. Therefore we did not feel that we analyzed or prioritized estuaries well for the CIT region.

Estuaries were treated in this portfolio several ways. Both the terrestrial ecological systems and the shoreline ecosystem targets identified saltmarsh communities. In both cases the definition of estuaries was associated with a vegetation type. Other estuaries were captured in the shoreline ecosystem categories as “mud flat,” “sand flat,” and “sand and gravel flat.” Here the definition was associated with gently sloping flats near river mouths represented as linear features.

The shore-zone data does not fully represent the spatial delineation of estuaries as polygons. Although the data does delineate some of the estuaries in the study region, many estuaries are left out of the data sets because they were not identified as the dominant feature in the intertidal. The strength of the shore-zone estuaries is their identification of substrate and vegetation attributes, not the spatial delineation.

The Canadian Wildlife Service and Environment Canada’s Pacific Estuary Conservation Program (PECP) delineated estuaries as polygons but with limited attribute information. This mapping made use of existing information (primarily TRIM, the 1:50,000 Watershed Atlas and the Marine Charts) and detailed specifications to direct the identification and digitizing of the intertidal zone and supratidal/upstream zone of each estuary. Aerial photographic interpretation and field work were not part of the methodology. The result is a comprehensive and standardized map of BC’s estuaries depicted as polygons. These data do not yet include the Fraser, Skeena, or Nass River estuaries since the methodology to delineate these large estuaries is still under development. The outer extent of each estuary (the intertidal zone) was delineated using the lowest normal tide line (zero chart contour line) from the Marine Charts. The inner extent (supratidal/upstream zone) was delineated by the TRIM coastline and 500 m upstream. Currently they are seeking data on the maximum upstream detection of surface salinity on these systems as an indicator of the upstream boundary of the estuary. Since these data focus on the delineation of estuaries and not the benthic and vegetation information within them, we did not use them in the nearshore analysis.

We therefore did not feel that the analysis strictly optimized for high priority estuaries as a coarse filter target. Until we have compiled comprehensive spatial and attribute information on estuaries, they will continue to be under-represented in site selection algorithms. As a result, estuaries and their importance were not emphasized enough in building the nearshore portfolio.

2.4.2 Offshore Marine Analysis

2.4.2.1 Overview

The CIT marine ecosystem spatial analysis (ESA) consists of 93 features, both biological and physical: considering representivity, distinctiveness, focal species, and rare or threatened species. Data were compiled from Fisheries and Oceans Canada (DFO), BC Ministry of Sustainable Resource Management (MSRM), Canadian Wildlife Service (CWS), Natural Resources Canada (NRCan), private researchers, and local knowledge (Table 2.24).



Page 102

Table 2.24 Summary of data layers by type

“Filter” Feature category Feature sub-category No. of layers

Coarse Filter Regional Representation Data Regions 6

Coarse Filter Ecosystem Representation Ecosections 8

Coarse Filter Ecosystem Representation Ecosystem Regions 3 regions + 3 sub-regions

Coarse Filter Ecosystem Representation Enduring Features & Processes 7 exposure + 21 substrate/depth

Fine Filter Focal Species Flora 13

Fine Filter Focal Species Seabirds 15

Fine Filter Focal Species Anadromous Spp. Richness x Stream Magnitudes

1

Fine Filter Focal Species Mammals 1

Fine Filter Focal Species Fish 1

Fine Filter Special Elements Rarity 6

Fine Filter Special Elements Distinctive Features 4 complexity + 4 current

TOTAL 48 Coarse Filter + 45 Fine Filter 93

In the following sections, each of these feature categories is discussed. For a more detailed table of the features, refer to Appendix 3.1.

2.4.2.2 Focal Species and Special Elements 2.4.2.2.1 Focal species

Focal species have received a lot of attention in terrestrial conservation (e.g., Noss 1991; Lambeck 1997), but have received less attention in marine conservation (e.g., Day and Roff 2000; Zacharias and Roff 2001; Roberts et al. 2003). Different categories of focal species exist, such as indicators, keystone, umbrella, and flagship species (for a complete discussion, see Zacharias and Roff 2001). A common concept in terrestrial conservation is that of the umbrella species, whose conservation is believed to also spatially protect other species’ habitat. Unfortunately, umbrella species are not as widely applicable in the marine environment, though they can prove valuable at more local scales (Zacharias and Roff 2001). One problem with the applicability of this concept to marine systems is that many candidate umbrella species, fitting the typical (terrestrial) apex predator profile, such as killer whales (Orcinus orca), exhibit massive migrations and use areas too large to be useful as marine umbrella species at most planning scales.

On the other hand, marine focal species can still be identified that are useful in conservation. Zacharias and Roff (2001) note that composition indicators, or species whose presence indicates other species or are used to characterize a particular habitat or community are particularly useful. They feel that sea birds, sea grasses, macroalgae, and benthic invertebrates are good candidates for focal species. We feel that sea birds may be also be seen at least partially as umbrella species, since protecting their foraging habitats will afford some protection to their prey species. Likewise, kelp beds (Nereocystis luetkeana and Macrocystis intergrifolia) were treated as local-scale umbrellas for the many species associated with them, as were eelgrass beds (Zostera sp.). Herring (Clupea



Page 103

pallasii) spawn were treated as a keystone species, since so many other species are attracted to, and rely upon, these areas to feed on the eggs (Hay and McCarter 2000).

Flora

For the CIT marine ESA, we considered the following focal vegetation species: Eelgrass, kelp, marsh grasses (Salicornia sp.), surf grasses (Phyllospadix sp), and a general shoreline vegetation class, aggregated from the BC Shorezone classification that includes Fucus, Ulva, halosaccion layers, “reds,” “soft browns,” and “chocolate browns.” (For a more detailed shoreline vegetation analysis, we deferred to the nearshore ESA team −see Section 2.4.1).

Seabirds

All major BC breeding seabird populations and colonies were considered: Ancient Murrelet, Black Oystercatcher, Cassin’s Auklet, Cormorant sp., Glaucous-winged Gull, Pigeon Guillemot, Puffin sp., Rhinoceros Auklet, and Storm Petrel sp. (data provided by Canadian Wildlife Service). In addition, very small islets, far from shore were also considered as surrogates for unsurveyed colonies (G. Kaiser, pers. comm., Map 25).

Seabirds are known to prefer certain marine waters. These we treated as “habitat capability” layers. We considered pelagic seabirds (shearwaters, fulmars, albatross, some gulls, and terns); waterfowl (ducks, swans, geese, grebes, and loons); and shorebirds (oystercatchers, sandpipers, plovers, and turnstones). Data were provided by Decision Support Services, Ministry of Sustainable Resource Management, based on known distributions and expert opinion.

Moulting sea ducks (Scoter sp. and Harlequin Ducks) inhabit certain nearshore BC waters during summer months. Because they are unable to fly, they are particularly susceptible to stressors such as oil spills (Savard 1988). These areas were also considered separately for each species grouping (data from CWS Coastal Waterbird Inventory; and from Savard 1988, digitized by J. Booth, Map 25).

Anadromous Streams

BC’s anadromous streams were captured using a species richness x stream magnitude ranking. Eight of BC’s nine anadromous spp were considered (eulachon, the ninth, was treated separately). These include all Oncorhynchus spp. and Dolly Varden (Salvelinus malma). About 1 out of 10 BC stream systems were considered likely to support significant numbers of anadromous species. Of those, about half were assigned a low score (1–4 out of a possible 24), meaning that they are small streams supporting only a few species. Only the Fraser River (outside the CIT study area) received a top score (24), with the Nass and Skeena rivers tied in second place (20). For a full description of this layer, refer to Appendix 3.2 Stream Richness x Magnitude (Map 27).

Steller Sea Lion

Steller Sea Lion (Eumetopias jubatus) haul-outs and rookeries were ranked on a scale of 1-4 based on population density.

Herring spawn

Herring spawn (Clupea pallasii) shorelines were ranked on a density measure based on DFO’s Spawn Habitat Index (Hay and McCarter 2001), using the latest available times series data (DFO



Page 104

2002). Data were cube root transformed and standardized to shoreline length per hexagonal planning unit.

2.4.2.2.2 Rare and threatened species

Rare, threatened and endangered species are generally given a lot of conservation attention. However, the inaccessible nature of the sea makes it much harder to survey and therefore know most of what is rare. Declining populations may go unnoticed through to their extirpation (Thorne-Miller 1999). In the marine ESA, we consider five Special Elements, on account of their rare or threatened status: hexactinellid sponge reefs, eulachon estuaries, sea otter (not WCVI), estuaries containing red or blue listed species, and Marbled Murrelet marine habitat.

Hexactinellid Sponge Reefs

Hexactinellid sponge reefs are unique to the BC coast and are important in terms of their ecology and their similarity to extinct Mesozoic sponge reefs. There is already evidence that they have been damaged by bottom trawling (Conway 1999; Conway et al. 2001; Krautter et al. 2001). In the spring of 2002, while setting a mooring to monitor one of the last undisturbed mounds, researchers discovered that it had been trawled since the previous visit (K. Conway, pers. comm., July 2002). We strongly support the recommendations of Conway (1999), Krautter et al. (2001), and Jamieson and Chew (2002), all who suggest that these sponge reefs be permanently protected from trawling. Since the summer of 2002 they have been given some protection in the form of a fishing closure, however closures can be lifted at any time at the discretion of fisheries managers. There are only four such reefs known to exist in the world, all of which are in the CIT study area.

Eulachon Estuaries

Eulachon (Thaleichthys pacificus) are an ecologically and culturally important fish species (Hart 1973). Eulachon spawning areas in the Central Coast are limited (McCarter and Hay 1999). Although larval eulachon spend very little time (hours) in their natal streams, the associated estuary or inlet is important juvenile habitat. Eulachon streams and estuaries should therefore be considered for protection.

Eulachon are heavily preyed upon during spawning migrations by spiny dogfish, sturgeon, Pacific halibut, whales, sea lions, and birds. In the ocean, it is also preyed on by salmon and other large fish (Pacific States Marine Fisheries Commission 1996; Fishbase 2001).

Data were downloaded from DFO Habitat and Enhancement Branch’s public Web site (DFO 2003), and were compared to FISS data, and published literature (McCarter and Hay 1999). Points were snapped to the BC Watershed Atlas when appropriate.

Sea Otter

Sea otters (Enhydra lutris) were once abundant throughout the Northeast Pacific but were hunted to near extinction from the mid-1700s to early 1900s. Apocryphally, the last known sea otter in British Columbia was accidentally shot in 1929. Between 1969 and 1972, 89 sea otters were reintroduced to Checleset Bay off northwest Vancouver Island and the population has been increasing at a rate of 17% per year (Estes 1990; Watson unpublished). Sea otters are important predators of invertebrates such as sea urchins and have been shown to play an important ecological roll as a keystone predator (Estes 1990).



Page 105

Unlike other marine mammals, sea otters do not have a blubber layer. They rely on their fur to keep warm and are therefore particularly vulnerable to oil spills, even minor ones. Several thousand (approx. 5,000) sea otters died in the 1989 Exxon oil spill in Valdez, Alaska (Marine Mammal Center 2000).

While the WCVI population appears to be increasing, the only known established colony in the CIT study area is in the Goose Islands.

Red-blue Estuaries

Estuaries in the North Coast and QCI harbouring provincially red (rare) or blue (threatened) listed species, mainly birds, were identified by Remington (1993), and digitized by Living Oceans Society for the CIT.

Marbled Murrelet Marine Habitat Capability

Marbled murrelets, in the auk family, are on the provincial “blue” list of vulnerable species. They may be moved to the “red” list of endangered species in the near future since the marbled murrelet population has suffered an estimated 40% drop in the past decade alone (Cannings and Cannings 1996). Both natural and human-related factors may be contributing to the species’ decline; potential causes include the loss of suitable nesting habitat, accidental death in gill-nets, oil pollution, increases in predator populations, and declines in food supplies due to recent El Nino events (SEI 1999).

Marbled murrelets lay a single egg on wide, mossy branch of old-growth conifer trees (Cannings and Cannings 1996). Therefore, during breeding season, murrelets can be found foraging just offshore of old-growth forests. Concentrations of foraging murrelets are sometimes found associated with tidal rips, high current areas, or river plumes. Researchers have identified a marbled murrelet juvenile nursery area in a semi-protected Nereocystis bed in Alaska (Kuletz and Piatt 1999). Although no similar areas have been identified in the Central Coast of BC, kelp beds and high current areas have also been considered in the marine ESA.

Marbled Murrelets are known to prefer certain marine waters. These we treated as a “habitat capability” layer. Data were provided by Decision Support Services, Sustainable Resource Management, based on known distributions and expert opinion.

Habitat-Forming Corals

We considered areas known to harbour large habitat-forming corals, which may well be threatened or endangered, but due to a lack of surveys their status largely remains unknown. Coral outcrops and “forests” are important habitat for adult fishes, crustaceans, sea stars, sea anemones and sponges because they provide protection from these currents and from predators. Some commercially important fish species are found in association with these reefs, such as Atka mackerel, Pleurogrammus monopterygius, and shortspine thornyhead, Sebastolobus alascanus, in Alaska. Rockfish are associated with Primnoa corals in the Gulf of Alaska (Etnoyer and Morgan 2003).

2.4.2.2.3 Distinctive features

One shortcoming of a representative areas approach is that it requires examining and possibly setting aside very large areas. Pragmatically, there may not be the political will or management



Page 106

capability to fully realize this approach. Furthermore, smaller but ecologically valuable areas may be passed over. Roff and Evans (2002 unpublished) argue that such smaller “distinct” areas are by definition different from their representative surroundings and may harbour higher (or lower) species diversity, richness, and abundance. These, they suggest, must also be considered in reserve design. Distinctive areas may also be thought of as representative of a certain type of habitat, but at a finer scale than the nominal scale of the study (John Roff, pers. comm.). In the marine CIT ESA, we included two separate indicators of distinctive habitats: Benthic topographical complexity, and high current.

Benthic Complexity

Areas of high taxonomic richness are often associated with areas of varying habitat. The more kinds of niches available in which organisms can live will usually lead to a wider variety of organisms taking up residence. Furthermore, the complexity of habitat can interrupt predator-prey relationships that in a simpler habitat might lead to the clear dominance or near extirpation of certain species (e.g., Eklov 1997). Thus, in complex habitats species may co-exist in greater diversity where elsewhere they might not. Likewise, a greater variety of life stages may also be supported. Thus, complex habitats may exhibit greater ecosystem resilience (e.g., Risser 1995; Peterson et al. 1998). Furthermore, if complex habitats do encourage biodiversity, as is being suggested, then it follows that they likely also offer greater resistance to invasive species (Kennedy et al. 2002).

Benthic topographical complexity is indicated by how often the slope of the sea bottom changes in a given area; that is, the density of the slope of slope of the depth. Note that this is not the same as relief, which looks at the maximum change in depth. Benthic complexity considers how convoluted the bottom is, not how steep or how rough, though these both play a role. Complexity is similar but not the same as “rugosity” as is sometimes used in underwater transect surveys, whereby a chain is laid down over the terrain and its length is divided by the straight-line distance. Rugosity can be strongly influenced by a single large change in depth, however, whereas complexity is less so, since all changes are treated more equally (Ardron 2002).

We used this analysis because we felt it captured biologically and physically meaningful features that the other measures missed. For example, archipelagos and rocky reefs are invariably picked out as areas of higher benthic complexity. Both are associated with several marine values. While “obvious” to the casual observer, they had hitherto no simple quantitative definition that could be used to identify them using a GIS. Benthic complexity will often also identify physical features such as sills, ledges, and other distinctive habitats that are associated as biological “hotspots” providing upwellings, mixing, and refugia (Ardron 2002, Map 29).

In the marine ESA, benthic complexity was examined separately within each of the four Ecological Regions (inlets, passages, shelf, slope).

High Current

This layer was extracted from the BC Marine Ecological Classification, version 2 (LUCO 1997, AXYS 2001), as well as incorporating additional local knowledge. High Current is defined as waters that regularly contain surface currents (tidal flow) greater than 3 knots (5.5 km/h or 1.5 m/s). These are areas of known mixing and distinctive species assemblages. In addition, high current areas often represent physical “bottlenecks” to water movement and as such are important to larval transfer and nutrient exchange.



Page 107

The strong currents of the southern half of the Central Coast, particularly in Johnstone Strait and Discovery Passage, are probably the most influential oceanographic variable of that region. They mix the water column so that nutrients, oxygen, temperature and salinity levels are almost uniform throughout (Thomson 1981). The constant re-suspension of nutrients in particular is most likely responsible for the rich biota of the south Central Coast passages. Mann and Lazier (1996) explain that tidally induced mixing in relatively shallow coastal waters prevents stratification of the water column, but the potentially adverse effects on phytoplankton are more than compensated for by the increased nutrient flux to the water column from the sediments. Annual primary productivity in tidally mixed areas tends to be above average for coastal waters (Mann and Lazier 1996). Highly productive and biologically diverse areas, such as the world-renowned dive site, Browning Passage (Queen Charlotte Strait), result from these nutrient-rich, mixed waters.

Because high current areas are always well mixed subsets of whatever larger mixing regime may exist, we have classified them as distinctive areas. They were considered separately for each of the four Ecological Regions (inlets, passages, shelf, slope).



Page 108

3.0 SETTING GOALS

3.1 Background

Explicit and quantitative goals are fundamental to systematic conservation planning (Margules and Pressey 2000). Goals represent the end toward which conservation efforts are directed for targeted species, communities, and ecosystems. Goals provide the quantitative basis for identifying and prioritizing areas that contribute to a network of conservation areas. Moreover, tracking progress toward goals provides an evaluation of the performance of a conservation program, from the scale of individual projects up to province-wide, national, and ultimately global programs. Establishing goals is among the most difficult—and most important—scientific tasks in conservation planning (e.g., How much protected area is enough? How many discrete populations and in what spatial distribution are needed for long-term viability?). As some have pointed out (e.g., Noss 1996; Soule and Sanjayan 1998), questions such as these cannot be answered satisfactorily by theory, but require an empirical approach, target-by-target, and a commitment to monitoring and continual re-evaluation over the long term.

Goals for conservation targets specify the number and spatial distribution of on-the-ground occurrences. As a general rule, a broad goal is to conserve multiple examples of each target, stratified across its geographic range in such a way that we capture the variability of the target and its environment. Replication of occurrences of each target must be sufficient to ensure persistence in the face of environmental stochasticity and the likely effects of climate change.

Although we were not able to rigorously analyze population viability of our focal species in this study, we have considered the conditions that contribute to viability in a general way through the use of a number of surrogates related to landscape condition and context (e.g., roadedness, landscape alteration). Conservation goals should support the evolutionary pathway of species in continually changing ecosystems, looking into the future at least 100 years or 10 generations. While that concept of viability could be said to apply to all targets, in practice we use several closely related, though distinct, groups of targets. Strategies for focal species emphasize recovery and evolutionary adaptation of individual species. In addition to species viability, coarse-filter strategies emphasize the conservation of ecosystem services (e.g., air, water, nutrient cycling, etc.), perhaps better characterized as ecological integrity at a regional scale (Pimentel et al. 2000). While conservation goals for species correctly emphasize potential genetic fitness and the functional roles of species in ecosystems, coarse-filter goals focus on representation of ecological variability and environmental gradients.

3.1.1 How Much Is Enough: Level of Landscape Protection (Representation Analysis)

Early work with mathematical site-selection algorithms concentrated on determining the “minimum set” conservation network, often defined as the area needed to protect at least one example (e.g., a individual plant, 1 occurrence of a mammal species or, in some cases, an entire population) of all selected elements of biodiversity (Rodrigues et al. 1999; Church et al. 2000; Reyers et al. 2000; Rodrigues and Gaston 2001). Most algorithms minimize the amount of area needed through complementary selections, or selecting areas based on the number of unique elements contained. This work has shown that, as biodiversity and endemism increase, so does the amount of area needed to represent all elements. In the regions with the highest biodiversity



Page 109

and many species with narrow distributions (i.e., tropical regions), Rodrigues and Gaston (2001) predicted that as much as 93% of the area would be needed to represent at least a single example of each biodiversity element, and that globally, 74% of global land area would need to be protected to conserve global biodiversity. Because northern temperate regions have lower biodiversity and fewer narrow endemics than found in the tropics, it is expected that less protected area is necessary to represent single occurrences of biodiversity elements. Nevertheless, Cumming et al. (1996) found that any reduction in total area of boreal forest resulted in the loss of forest structural diversity across the landscape, and that forest structure is critically important for supporting biodiversity in boreal forests. Importantly, site selection algorithms, by themselves, do not address the more difficult and real-world questions concerning the area needed to maintain viable populations of species and the persistence of biodiversity. Undoubtedly, maintaining viable populations requires vastly more area than representing single elements, particularly if those single elements are defined as single occurrences of individuals.

An intended application of conservation science is to assist in answering these more difficult questions, through explicitly defining the area and configuration of habitats needed for the long-term persistence of biodiversity. Many studies have proposed minimum targets for biodiversity conservation, either generally or for specific regions. In some cases these figures are based on estimates by experts of the area necessary to maintain viable populations, ecosystem services, or the persistence of biodiversity generally; in other cases they are based on the empirical results of studies or population viability analyses in combination with site-selection algorithms (Table 3.1). The implicit objective of these recommendations is to reduce extinction rates to near-background levels, maintaining the integrity of all ecosystems, and to sustain natural ecological flows and processes on a regional scale. When existing protected areas—which generally were not selected on the basis of biological criteria—are included in designs, the results are less efficient, and more land (e.g., 70% of the Greater Yellowstone Ecosystem, where 27% of the landscape is already protected; Noss et al. 2002) is needed to meet similar conservation goals. Using spatially explicit population models linked to site selection procedures, Carroll et al. (2003) determined that at least 37% of their U.S.–Canadian Rocky Mountain study area would need to be protected to meet population viability criteria for large carnivores (grizzly bear and wolf). Their modelling procedures preferentially selected the most productive (e.g., source) habitats, based on estimated fecundity, mortality, and connectivity parameters. In other planning efforts, without spatially explicit population modelling, it may be impossible to select, a priori, similar critical (i.e., irreplaceable) sites in optimal configuration, and the precautionary principle dictates that higher levels of targets should be set.

3.1.2 Goals for Focal Species

Large carnivores are often selected as focal species, given their often low demographic productivity, large area requirements, low density, and potential sensitivity to landscape change (Noss et al. 1996). In many areas of western Canada and northwestern U.S., large predators, such as the grizzly bear, wolf, or wolverine, are used as surrogates for biodiversity in the development of conservation areas or protected area designs. These species require large contiguous or linked areas of high-quality habitats to ensure viable populations.

For such species, population viability should be assessed to determine how many individuals are necessary to insure a high probability of long-term survival. While estimating population



Page 110

viability has been proposed as a major objective of conservation science, the necessary data required to accurately determine the viability of populations are usually absent ( Shaffer 1981; Boyce 1992; Morris and Doak 2002). These data include vital demographic rates (e.g., age-specific mortality rates, mean litter size, sex ratio, inter-birth intervals) that influence population growth and decline, as well as the natural level of variance in all vital rates. Determining vital rates is costly and time consuming, and can take years of intensive study. Consequently, most attempts to assess population viability and subsequently determine “how much is enough” often result in a correct and responsible conclusion of uncertainty, especially when natural environmental fluctuations are also factored into predictive estimates. Even when vital rates are known (or estimated), the exact linkages between habitat area/habitat quality requirements and vital demographic rates are often not well enough understood to draw conclusions with any confidence regarding “how much is enough.”

3.1.3 Conclusions

In this study, we did not have the time or funding to gather the necessary data and perform detailed, spatially explicit population modeling for the selected focal species. Therefore, we were unable to evaluate the potential population viability of these species in alternative networks of reserves compared to the current network of protected areas. We recommend that further studies be conducted to gain a more detailed understanding of how much area, and in what configuration and type of management, is needed for persistence of focal species over time.

Although it is difficult to directly transform the knowledge available from other studies into specific management prescriptions for our study region, the research reviewed above does provide insight into the question of “how much is enough” (see Table 3.1). The percentages cited in Table 3.1 generally call for between 25 and 75% of a region to be protected in order to meet stated conservation goals relating to ensuring the long-term viability of conservation targets. It is important to recognize that the amount of land needed to meet similar goals will vary considerably by region, depending on the physical and biological heterogeneity of the region, habitat quality, human land uses, and many other factors (Noss and Cooperrider 1994). The proportion of the landscape provided with protection, and the form and function of that protection also will vary across regions, depending upon social, political, and economic constraints. Higher proportions of the landscape devoted to protection could certainly be argued for, particularly given the sparse data on the ecological dynamics and requirements to maintain focal species and natural processes. Indeed, the precautionary principle dictates higher levels of protection to buffer against these uncertainties.

Table 3.1 Percentage of land recommended for protection in a number of regions, based either on general estimates of viability requirements or on empirical results of goal-driven site-selection algorithms or other quantitative studies

Source Region Goal Recommended Area

Odum (1970) Georgia Optimize ecosystem services and quality of life in a self-sustaining ecosystem

40%



Page 111

Source Region Goal Recommended Area

Odum and Odum (1972) General

Optimize ecosystem services and economic and cultural well-being

50%

Noss (1993)

Oregon Coast Represent all plant species and wetland types at least once

50%

Cox et al. (1994) Florida Represent all bird, mammal, and plant species at least once

33.3%

Mosquin et al. (1995)

Canada Represent all bird, mammal, reptile, and plant species at least once

35%

Ryti (1992)

San Diego Canyons Maintain an effective population of 500 grizzly bears (total pop. = 2000)

65%

Ryti (1992)

Islands in Gulf of California Protect all clusters of rare species and community occurrences and all primary forest; provide for carnivore recovery

99.7%

Margules et al. (1988)

Australian river valleys Protect rare species and natural communities

44.9% - 75.3%

Noss (1996)

General Meet well accepted conservation goals in various regions

25% – 75%

Noss et al. (1999)

Klamath-Siskiyou Protect roadless areas that meet all special elements, representation, and focal species goals

60% – 65%

Hoctor et al. (2000)

Florida Capture biological priority areas and provide connectivity statewide

50%

Rodrigues & Gaston (2001)

Tropical region Represent each species at least once

93%

Rodrigues & Gaston (2001)

Globally Represent each plant (and vertebrate) species at least once

74%

Noss et al. (2002) Greater Yellowstone Ecosystem

Protect megasites that meet all special elements, representation, and focal species goals

43%

Carrol et al. (2003) US-Canada Rocky Mnts Protect highest-quality habitat and source areas to maintain viable populations of carnivores

37%

Cowling et al. (2003)

Cape Floristic Region, South Africa

Capture a comprehensive set of targets representing biodiversity patterns and processes

52%

The Nature Conservancy (2003, unpublished)

> 50 ecoregional assessments in 11 countries

Meet multiple conservation goals, especially representation of ecosystems and protection of special elements

Mean: 30-40%

(up to ca. 70%)



Page 112

3.2 CIT ESA Terrestrial/Freshwater Goal Setting

Given the time and resources available to the ESA, the most advisable approach to setting goals draws on the EBM framework by setting a range of goals that can be used to construct separate portfolios for several goal levels within that range. Using this approach, a series of 5 potential conservation solutions were created designed to represent 30%, 40%, 50%, 60%, and 70% of total available terrestrial ecosystem area, freshwater ecosystem stream length, and focal species habitat values. These solutions were then used for purposes of prioritization, allowing the team to compare areas that were necessary to satisfy all goal scenarios including the minimal goal set, to areas only selected in larger goal sets, to those areas never selected, regardless of the goal setting.

3.3 Goals for the Marine Nearshore Environment

The marine team set goals for each shoreline ecosystem and species target, with some additional advice from regional experts. We set general guidelines for establishing goals for marine targets, expressed as percentages of the “amount” of the target data (i.e., 30% of all spatially delineated herring spawn sites represented as lineal metres) throughout the ecoregion. The purpose of these guidelines was to set conservation goals that address the rarity of the target, the number of occurrences, and its distribution across the ecoregion.

There was no mechanism for us to consistently assess viability for fine and coarse filter marine targets. Since we couldn’t rely on the marine data sets to provide information about the relative viability of individual occurrences, we set conservative (low) goals that would help drive the algorithm to assemble an efficient portfolio around the sites most important to multiple targets. We therefore attempted to answer the question “where do we start?” in conserving places for nearshore biodiversity, as opposed to “how much (area) is enough?” We based the goals on importance of the targets, co-occurrence of habitat types and species, least cost (or the most suitable places), and our confidence in individual data sets.

3.3.1 Representation (Coarse Filter) Goals

The shore-zone data were the most uniform across the ecoregion and provided us with the best data for describing a portfolio that is representative of nearshore marine biodiversity. We experimented with two different goal scenarios and compiled the results from the “summed solution” outputs (the number of times a planning unit is chosen out of the total number of iterative runs in a scenario) to a single gradient. Goals for individual shoreline ecosystem targets were 10% and 20% (Table 3.2). Lacking further ecological justification to determine the relative importance of individual shoreline ecosystems, we set goal for all ecosystems the same. In the same light, coarse filter intertidal vegetation and habitat targets were run at goals of 15% and 30%, respectively (Table 3.3).

We acknowledge that 10% and 20% of the existing extent of those ecosystems is not the same as 10% or 20% of historic composition, but we currently lack data to consistently define the historic extent of shoreline and estuary ecosystems throughout the study area.



Page 113

Table 3.2 Shoreline ecosystem goals

Target Target_ID Shore units

Total length (m)

10% goal

10% goal amount

20% Goal

20% Goal Amount

Channel 1001 3 1,548.5 0.1 154.9 0.2 309.7

ChannelE 1002 9 2,690.1 0.1 269.0 0.2 538.0

ChannelP 1003 73 38,751.9 0.1 3,875.2 0.2 7,750.4

Estuary Wetland 1004 40 43,949.0 0.1 4,394.9 0.2 8,789.8

Estuary WetlandE 1005 13 9,942.3 0.1 994.2 0.2 1,988.5

Estuary WetlandP 1006 1134 1,046,811.2 0.1 104,681.1 0.2 209,362.2

Estuary WetlandVE 1007 9 9,323.4 0.1 932.3 0.2 1,864.7

Estuary WetlandVP 1008 47 46,032.7 0.1 4,603.3 0.2 9,206.5

Gravel Beach 1009 215 78,725.4 0.1 7,872.5 0.2 15,745.1

Gravel BeachE 1010 333 93,229.1 0.1 9,322.9 0.2 18,645.8

Gravel BeachP 1011 2983 941,273.4 0.1 94,127.3 0.2 188,254.7

Gravel BeachVP 1012 12 4,336.2 0.1 433.6 0.2 867.2

Gravel Flats 1013 28 15,052.7 0.1 1,505.3 0.2 3,010.5

Gravel FlatsE 1014 56 19,683.4 0.1 1,968.3 0.2 3,936.7

Gravel FlatsP 1015 148 76,247.2 0.1 7,624.7 0.2 15,249.4

Gravel FlatsVE 1016 9 3,599.3 0.1 359.9 0.2 719.9

High Tide Lagoon 1017 50 85,740.2 0.1 8,574.0 0.2 17,148.0

High Tide LagoonE 1018 10 4,254.7 0.1 425.5 0.2 850.9

High Tide LagoonP 1019 210 236,286.4 0.1 23,628.6 0.2 47,257.3

High Tide LagoonVP 1020 60 80,393.8 0.1 8,039.4 0.2 16,078.8

Man-made (not a target) 1021 13 5,253.3 0 0.0 0 0.0

Man-madeE (not a target) 1022 5 1,262.8 0 0.0 0 0.0

Man-madeP (not a target) 1023 302 135,230.8 0 0.0 0 0.0

Mud Flat 1024 7 4,985.3 0.1 498.5 0.2 997.1

Mud FlatP 1025 156 107,010.0 0.1 10,701.0 0.2 21,402.0



Page 114


Total length (m)

10% goal

10% goal amount

20% Goal

20% Goal Amount

Rock Cliff 1026 1396 703,588.4 0.1 70,358.8 0.2 140,717.7

Rock CliffE 1027 4314 1,768,132.6 0.1 176,813.3 0.2 353,626.5

Rock CliffP 1028 11822 5,833,450.0 0.1 583,345.0 0.2 1,166,690.0

Rock CliffVE 1029 218 125,300.5 0.1 12,530.1 0.2 25,060.1

Rock CliffVP 1030 68 31,122.4 0.1 3,112.2 0.2 6,224.5

Rock Platform 1031 124 33,573.1 0.1 3,357.3 0.2 6,714.6

Rock PlatformE 1032 1774 661,917.3 0.1 66,191.7 0.2 132,383.5

Rock PlatformP 1033 447 128,569.6 0.1 12,857.0 0.2 25,713.9

Rock PlatformVE 1034 345 162,097.7 0.1 16,209.8 0.2 32,419.5

Rock with Gravel Beach 1035 710 287,587.8 0.1 28,758.8 0.2 57,517.6

Rock with Gravel BeachE 1036 3374 1,394,014.4 0.1 139,401.4 0.2 278,802.9

Rock with Gravel BeachP 1037 9679 4,406,751.8 0.1 440,675.2 0.2 881,350.4

Rock with Gravel BeachVE 1038 123 40,373.1 0.1 4,037.3 0.2 8,074.6

Rock with Gravel BeachVP 1039 74 35,028.4 0.1 3,502.8 0.2 7,005.7

Rock with Sand & Gravel Beach 1040 173 57,913.7 0.1 5,791.4 0.2 11,582.7

Rock with Sand & Gravel BeachE 1041 2333 876,653.2 0.1 87,665.3 0.2 175,330.6

Rock with Sand & Gravel BeachP 1042 7600 2,843,714.0 0.1 284,371.4 0.2 568,742.8

Rock with Sand & Gravel BeachVE 1043 12 11,238.0 0.1 1,123.8 0.2 2,247.6

Rock with Sand & Gravel BeachVP 1044 39 15,471.3 0.1 1,547.1 0.2 3,094.3

Rock with Sand Beach 1045 19 4,346.8 0.1 434.7 0.2 869.4

Rock with Sand BeachE 1046 363 124,830.9 0.1 12,483.1 0.2 24,966.2

Rock with Sand BeachP 1047 544 182,404.5 0.1 18,240.5 0.2 36,480.9

Sand & Gravel Beach 1048 173 59,406.9 0.1 5,940.7 0.2 11,881.4

Sand & Gravel BeachE 1049 198 53,307.5 0.1 5,330.7 0.2 10,661.5

Sand & Gravel BeachP 1050 3600 1,339,292.9 0.1 133,929.3 0.2 267,858.6

Sand & Gravel BeachVP 1051 29 8,876.9 0.1 887.7 0.2 1,775.4

Sand & Gravel Flat 1052 96 38,497.7 0.1 3,849.8 0.2 7,699.5



Page 115


Total length (m)

10% goal

10% goal amount

20% Goal

20% Goal Amount

Sand & Gravel FlatE 1053 176 99,267.0 0.1 9,926.7 0.2 19,853.4

Sand & Gravel FlatP 1054 2689 1,298,001.0 0.1 129,800.1 0.2 259,600.2

Sand & Gravel FlatVP 1055 29 12,829.3 0.1 1,282.9 0.2 2,565.9

Sand Beach 1056 19 10,393.3 0.1 1,039.3 0.2 2,078.7

Sand BeachE 1057 242 187,660.8 0.1 18,766.1 0.2 37,532.2

Sand BeachP 1058 298 124,058.7 0.1 12,405.9 0.2 24,811.7

Sand Flat 1059 42 23,653.5 0.1 2,365.4 0.2 4,730.7

Sand FlatE 1060 129 78,238.5 0.1 7,823.8 0.2 15,647.7

Sand FlatP 1061 2179 1,306,395.7 0.1 130,639.6 0.2 261,279.1

Sand FlatVP 1062 40 22,440.9 0.1 2,244.1 0.2 4,488.2

Undefined (not a target) 9999 1026 662,992.3 0 0.0 0 0.0

59 TARGETS 61,095 27,340,266.0 10% 2,734,026.6 20% 5,468,053.2

Total shoreline contains 62,441 units with 28,145,005.1 m

E = EXPOSED

P = PROTECTED

VE = VERY EXPOSED

VP = VERY PROTECTED




Page 116

Table 3.3 Vegetation and fine filter goals

Target Target_ID Total Length

(m) 15% Goal 15% Goal Amount

(m) 30% Goal 30% Goal Amount

(m)

Habitat 3 1063 3,674,186.6 0.15 551,128.0 0.3 1,102,256.0

Eelgrass 1064 2,780,418.1 0.15 417,062.7 0.3 834,125.4

Surfgrass 1065 2,092,980.4 0.15 313,947.1 0.3 627,894.1

Saltmarsh 1066 2,669,922.6 0.15 400,488.4 0.3 800,976.8

Kelp 1067 5,134,530.6 0.15 770,179.6 0.3 1,540,359.2

Herring Spawning 1068 418,717.1 0.15 62,807.6 0.3 125,615.1

Thick-billed Murre seabird colony 1075 8,654.3 0.15 1,298.1 0.3 2,596.3

Cassin’s Auklet seabird colony 1074 322,925.4 0.15 48,438.8 0.3 96,877.6

Common Murre seabird colony 1073 18,896.5 0.15 2,834.5 0.3 5,669.0

Ancient Murrelet seabird colony 1072 251,716.0 0.15 37,757.4 0.3 75,514.8

Brandt’s Cormorant seabird colony 1071 5,296.6 0.15 794.5 0.3 1,589.0

Horned Puffin seabird colony 1070 15,476.0 0.15 2,321.4 0.3 4,642.8

Tufted Puffin seabird colony 1069 99,504.6 0.15 14,925.7 0.3 29,851.4

All Vegetation and Fine Filter targets 17,493,224.8 0.15 2,623,983.7 0.3 5,247,967.4



Page 117

3.4 Goals for Offshore Marine

Halpern (2003) reviewed 89 studies of no-take marine reserves and found that regardless of size, marine reserves lead to increases in density, biomass, individual size, and diversity in all functional groups. However, larger reserves did produce larger increases. Halpern goes on to caution, “…that to supply fisheries adequately and to sustain viable populations of diverse groups of organisms, it is likely that at least some large reserves will be needed” (Ibid pp. 129–130).

A variety of Marine reserve sizes ranging from 10% to 50% have been suggested as being efficacious as a conservation and/or fisheries management tool (Carr and Reed 1993; Ballantine 1997; NRC 2000; Roberts and Hawkins 2000; MRWG 2001), with an emphasis on larger reserves coming from the more recent literature. Furthermore, it has been found that larger reserves often have beneficial effects disproportionate to their size (Halpern 2003). In the marine CIT ecosystem spatial analysis, we explored a variety of conservation goals (also know as “targets” in the literature) that produced overall areas ranging from 5 to 50% of the study area. Specifically, we looked at Marxan solutions that comprised 5, 10, 20, 30, 40, and 50% of the study area. However, this does not imply that equal amounts of each of our 93 feature elements were represented. Rather, as explained below, each feature was assigned a goal based on a range of six relative rankings.

Before choosing actual percentages per feature as a goal, we examined each data set and assigned to it a relative term, where “moderate” was taken as the common baseline or average value. The five terms used were: low, moderate-low, moderate, moderate-high, high, and very-high. In general, we assigned lower rankings such as low or moderate-low to features that were common (i.e., plentiful), and higher rankings features that were more unusual or rare. Umbrella and keystone species were generally assigned a moderate-high ranking. By using these six simple qualitative rankings, we were able to class the features relative to each other. Once that was completed, we could then implement a range of actual numerical targets and observe the effects. Such a strategy (though not in the context of MARXAN) has been suggested by Levings and Jamieson (1999) as “dimensionless scores,” to be used to meet various criteria such as distinctiveness, and naturalness. The addition of the computer software allows for quick feedback to compare scenarios. Table 3.4 displays the actual percentages attached to each qualitative ranking. Columns display each conservation scenario, while the rows display the rankings. Appendix 3.1 lists all 93 features in the marine ESA, and their assigned relative goals.

Table 3.4 Actual percentages attached to each qualitative ranking

Relative ranking Conservation goals (%)

Low 2 4 8 12 16 20

Mod-Low 4 8 16 24 32 40

Moderate 6 12 24 36 48 60

Mod-High 8 16 32 48 64 80

High 10 20 40 60 80 100

V. High 12 24 48 72 96 120*

Overall size 5 10 20 30 40 50

*Goals greater than 100% cannot be met, but do serve to give these features a higher emphasis.



Page 118

4.0 HUMAN IMPACTS

4.1 Background and Rationale

The biological integrity of terrestrial, nearshore marine and freshwater ecosystems depends largely on previous human alterations.11 Although it has been well established, through experimental and correlative studies, that ecological systems are adversely affected by human alterations, there is relatively little information about the functional relationship between ecological integrity, key ecological processes and human activities (Jungwirth et al. 2002). This lack of information hampers assessment of ecological integrity; consequently, standard methods for quantifying the relative degree of impacts and methods for using impact measures to prioritize areas for conservation have not been developed. Nevertheless, a number of studies have been completed that can provide direction with setting thresholds associated with ecological integrity (Table 4.1).

Watersheds defined both the unit of analysis and recommended management. There are multiple studies that suggest conservation action and management should take place at the scale of entire watersheds (Sullivan et al. 1987; Sheldon 1988; Williams et al. 1989; Moyle 1991; Naiman et al. 1992; Stanford and Ward 1992; Naiman et al. 1993, 2000; Pringle 2001; Baron et al. 2002). For example, many of the species and trophic systems of coastal B.C. (e.g., salmon spawning and rearing and the interactions between wildlife species and salmon) tend to be strongly linked to key ecological processes at a watershed-scale such as sedimentation control, regulation of flow regimes and nutrient cycling.

In addition, field studies suggest that watersheds are the appropriate scale to measure and manage cumulative human impacts. Measurable indicators tend to correlate with human activity data when measured at watershed scales, while the correlation is often absent at local scales (Karr 1991; Roth 1996; Muhar and Jungwirth 1998; Thorton 2000; Carignan et al. 2002; Pess et al. 2002). Thus, because watersheds define an appropriate ecological unit where human impacts tend to accumulate and can be measured and because of their value for key ecological processes and their global rarity, identifying and representing a range of intact watersheds should be included as a part of any credible, systematic, science-based conservation analysis.

Here we report methods for assessing relative impacts at multiple scales, using watersheds as our analysis unit, based on known linkages between human impacts and ecological processes. We report here simple evaluation criteria specifically designed to use surrogates for ecological integrity. We chose surrogates that 1) are likely to correlate with key ecological functions and processes found in intact ecosystems, 2) are measurable and mappable and 3) have region-wide data available with relatively uniform quality and coverage. In addition, we define standard comparison units, based on systematic and repeatable criteria for defining watershed boundaries based. We suggest that these tools, in combination, can be used to assess ecological integrity multiple scales, from third order watershed-scale to a regional scale and can thus provide critical information for systematic conservation planning efforts.

11 Biological integrity includes maintenance of both species viability (e.g., genetic fitness and the functional role of species in ecosystems) and ecosystem services (e.g., air, water, nutrient cycling), at a regional scale (Pimentel et al. 2000).



Page 119

4.2 Methods

We used a systematic set of decision rules to define watershed boundaries. We used the B.C. watershed atlas as the basis, since it provides established and documented spatial data. We defined additional units based on aggregating BC watershed atlas “third order” (i.e., LWSD) polygons into discrete units. Primary watersheds were created by grouping all watershed polygons that share a common saltwater exit point. Although primary watersheds define drainages, their size range covers several orders of magnitude (i.e., from less than 1 ha to over 5 million ha). Note that primary watersheds define an objective unit and can be sub-divided using any number of arbitrary methods. Therefore, to systematically classify sub-primary watersheds, we also defined two units: large river systems and intermediate river systems. These are sub-primary watersheds that are between 10,000 and 100,000 ha and 100,000 to 1,000,000 ha, respectively, and defined using a standardized set of decision rules (see Appendix 4.1 for details). This allows assessment of impacts and other characteristics at multiple spatial scales.

Using these criteria, we applied a scheme to assess watershed condition, based on a modified Moore (1991) methodology. Human impacted area was calculated by combining all human altered areas (clearcut, urban, agriculture) with a 200 m buffer area around roads. We used a 200 m buffer as a compromise between indirect impacts of roads on vertebrate species (i.e., “zone of influence”) which has been reported to range from 200 m to several kilometres and direct impacts of roads on adjacent habitat, which ranges from 20 to over 200 m (see Table 4.1 for summary). Overlapping areas were treated as impacted (i.e., overlaps were only counted once), which also allow calculation of overall percentage of development in any watershed unit. This method has several advantages including correcting for patchy data (e.g., where either logging data or road data is absent). Because some areas have relatively little vegetated areas and, consequently, little developable area and little productive habitat, we calculated impacted area as a percentage of potential vegetated area, which was calculated as a sum of natural vegetated area, human altered vegetated area and urban area. Road density was calculated as kilometre of road per square kilometre, also using potential vegetated area.

Watersheds with more than 2% of their area affected may still be considered ecologically intact, depending largely on both the cumulative impact of human alteration and the spatial location of human alterations. To identify such watersheds, we used two additional factors for assessing the overall level of impact, 1) proximity of impacts to rivers and streams and 2) road density. This allowed separation of moderately impacted areas from those with higher levels of human impacts (Table 4.2).

We also sought to identify relatively intact watersheds at multiple spatial scales. Small intact watersheds may be sufficient for harboring viable occurrences of some non-vagile species (e.g., rare plant communities), but larger, contiguous intact areas and characteristics present only in larger river systems are necessary to conserve viable populations of vertebrate species. We applied definitions based on Table 4.2 to several scales of watersheds and river systems (Table 4.3).

4.3 Results

Map 31 summarizes the impacts assessment for the study area according to the three broad classes of intact, modified, and developed.



Page 120

Table 4.1 Reported thresholds for human impacts on biodiversity

Report Results

Mace et al. 1996

High female Grizzly bear habitat use areas (i.e., within composite home ranges) had less than 0.6 km/1 km2 road density; comparable areas outside of composite home ranges had > 1 km/km2 road density.

McGurk and Fong 1995 Detrimental effects on aquatic ecosystems based on macro-invertebrate distribution, where roads cover >5% or more of a watershed

Mech 1989 0.6 km/km2 road density threshold for wolves

Van Dyke et al. 1986 0.6 km/km2 road density threshold for mountain lions

Forman et al. 1997 0.6 km/km2 road density threshold for grizzly bears

Findlay and Houlahan 1997 Species richness in Ontario, Canada wetlands was negatively correlated with proximity to roads at distances up to 1–2 km

Jones and Grant 1996; Jones et al. 2000

Partial logging (25% of watershed) significantly increases flood event magnitude. No measurable difference above 25% threshold (i.e., 100% logged areas had similar effects as 25%)

Schuler 1995 10% threshold for aquatic system permeability

Quarles et al. 1974; Dales and Freedman 1982

Soil contamination decreases exponentially away from roads; thresholds vary between 20 and 200 m

Lyon 1983; Paquet and Callaghan 1996; Rost and Bailey 1979

Elk and other large ungulate avoidance 100–200 m distance from road.

Forman 1995 Indirect impacts for wildlife (i.e., increased human access, mortality) range from 200 to 1000 m from a road

Table 4.2 ESA human impact classes. Areas without vegetative cover data are omitted from this analysis.

Class Description

Intact1 Pristine, no industrial impact

Intact2 Modified, < 2% area impacted and < 0.35 km/km2 road density

Intact3 < 10% of area impacted and < 10% of area in proximity to rivers/streams impacted and < 0.35 km/km2 road density

Modified1 < 15% of area impacted and < 0.6 km/km2 road density

Modified2 < 25% of area impacted and < 0.6 km/km2

Developed > 25% of area impacted or > 0.6 km/km2 road density



Page 121

Table 4.3 Watersheds and river systems

Classification Label

Primary watersheds

< 10,000 ha Small primary

10,000–100,000 ha Medium primary

100,000–1,000,000 ha Large primary

> 1,000,000 ha Very large primary

River systems

10,000–100,000 ha Intermediate river systems

100,000–1,000,000 ha Large river systems

Third order watersheds

LWSD polygons LWSD systems [note that these do not occur in Alaska]



Page 122

5.0 SPATIAL ANALYSIS: METHODS

5.1 Background

Our ecosystem spatial analysis is designed to provide an information base and decision support for subsequent planning and management efforts designed to address four well-accepted goals of conservation (Noss and Cooperrider 1994): 1) represent ecosystems across a range of environmental gradients; 2) maintain viable populations of native species; 3) sustain ecological and evolutionary processes; and 4) build a conservation network that is resilient to environmental change. In pursuit of these goals we integrate three basic approaches to conservation planning: 1) protection of special elements, including imperiled species, natural communities, and genetic variants; 2) representation of a broad spectrum of environmental variation (e.g., vegetation, geoclimatic, and aquatic classes); and 3) protection of habitats of focal species (Lambeck 1997; Miller et al. 1998/99), whose needs help planners address issues of habitat area, configuration, and quality. Together, these three tracks constitute a comprehensive approach to regional scale conservation planning (Noss et al.1999).

5.2 Spatial Analysis: Design Process and Tools

5.2.1 Steps of Spatial Analysis

For the CIT ESA, the challenge is to take an analysis of special elements, ecosystem representation, and focal species, and create spatially explicit assessments of where the region’s biodiversity values are located and what condition they are in. This information can then be used to inform ecosystem-based management of the region, including the placement of protected areas. To perform this assessment, the three-track approach is applied to freshwater, terrestrial, and marine environments via the following key steps:

1. Select conservation targets (e.g., special elements, focal species, and ecological systems) that will be used to characterize the biodiversity values within the study area.

2. Collect data for special element occurrences, develop focal species habitat models, and create ecosystem classifications that can be used to map the distribution of targets within the study area.

3. Using available data and models assess viability of targets.

4. Set conservation goals to serve as benchmarks for identifying conservation priorities and as an initial hypothesis about the level of effort required to conserve biodiversity.

5. Integrate information for special elements, ecosystem representation, and focal species in each of freshwater, terrestrial, and marine environments to create a spatially explicit assessment of conservation values for the study area.

This type of rigorous analysis employs thousands of pieces of detailed information. It requires location-specific information for conservation targets as well as the past, current, and potential future status of lands where they occur. The team used the best available information for this assessment but recognizes that new and more comprehensive data will continually become



Page 123

available and that ultimately, the ESA must be regarded as being a first step in an iterative assessment process.

5.2.2 Automated Site Selection Algorithms 5.2.2.1 SITES

Early conservation assessments and reserve designs depended on manual mapping to delineate sites and on simple scoring procedures to compare and prioritize sites. The large number of conservation targets and the large size and diverse types of data sets describing the targets in this study required the use of a more systematic and efficient site selection procedure. We used the site selection software SITES (v1.0), developed at the University of California, Santa Barbara under contract to TNC, as an aid to portfolio assembly. SITES operates within ArcView GIS as “an analytical toolbox for designing ecoregional conservation portfolios” (Andelman et al. 1999). SITES has been or is being used as an aid for designing and analyzing alternative portfolios in a number of TNC ecoregional plans, including the Northern Gulf of Mexico (Beck et al. 2000), Cook Inlet, Klamath Mountains, Sierra Nevada, Middle Rocky Mountains-Blue Mountains, and Southern Rocky Mountains ecoregions.

SITES uses an algorithm called “simulated annealing with iterative improvement” as a heuristic method for efficiently selecting regionally representative sets of areas for biodiversity conservation (Pressey et al. 1996; Csuti et al. 1997; Possingham et al. 1999). It is not guaranteed to find an optimal solution, which is prohibitive in computer time for large, complex data sets such as ours. Rather, the algorithm attempts to minimize portfolio “cost” while maximizing attainment of conservation goals in a compact set of sites. This set of objectives constitutes the “Objective Cost function:”

Cost = Area + Species Penalty + Boundary Length

where Cost is the objective (to be minimized), Area is the number of hectares in all planning units selected for the portfolio, Species Penalty is a cost imposed for failing to meet target goals, and Boundary Length is a cost determined by the total boundary length of the portfolio.

SITES attempts to minimize total portfolio cost by selecting the fewest planning units and smallest overall area needed to meet as many target goals as possible, and by selecting planning units that are clustered together rather than dispersed (thus reducing boundary length). SITES accomplishes this task by changing the planning units selected and re-evaluating the Cost function through multiple iterations. We had SITES perform 1,000,000 iterative attempts to find the minimum cost solution per simulated annealing run and perform 20 such runs for each alternative conservation scenario we explored. Alternative scenarios were evaluated by varying the inputs to the Cost function. For example, the Boundary Length cost factor was increased or decreased depending on the assumed importance of a spatially compact portfolio of sites, and a range of goals were used.

We used numerous SITES runs to determine alternative portfolios, which met, stated goals for protection of the target groups: vegetative, abiotic, and aquatic habitat types within the representation track; and high-quality habitat for the several species analyzed within the focal species track. We made SITES runs with and without existing and potential protected areas “locked in” to the portfolio, looking for differences in the location and area of selected planning



Page 124

units. While site selection algorithms have been used frequently in the past to select a single “optimized” conservation portfolio, the ESA team elected to use SITES in a manner that reflected the variability in potential conservation solutions. Using “summed runs,” the team generated a range of potential conservation portfolios for the purposes of identifying core areas—areas common to most or all outputs from SITES. This approach is meant to provide makers-makers with information about high value conservation areas while acknowledging that competing interests at LRMP tables may not allow for the implementation of any single reserve design emerging from the SITES algorithm.

5.2.2.2 MARXAN

Evolving out of SITES, MARXAN is a site optimization algorithm that has been used in several of the CIT ESAs. It was developed by Dr Hugh Possingham, University of Queensland, and Dr Ian Ball, now at Australian Antarctic Division in Tasmania. MARXAN comes from a lineage of successful selection algorithms, beginning with SIMAN, SPEXAN, and SITES. MARXAN was developed from SPEXAN and SITES in part to aid in work on the Great Barrier Reef Marine Park Authority’s re-evaluation of their park designations. To design an optimal reserve network, both MARXAN and SITES examine each individual planning unit for the values it contains. They then select a collection of these units to meet the conservation targets that have been assigned. The algorithm will add and remove planning units in an attempt to improve the efficiency of the reserves. What makes these algorithms different from other iterative approaches is that there is a random element programmed into them such that early on in the process the algorithm is quite irrational in what it chooses to keep or discard, often breaking the rules of what makes a good selection. This random factor allows the algorithm to choose less than optimal planning units earlier that may allow for better choices later. As the program progresses, the computer behaves more predictably—but not entirely. The process continues, with the criteria for a good selection getting progressively stricter, until finally the reserve network is built.

Given a sufficiently diverse set of features, it follows that because of the random element, no two runs are likely to produce exactly the same results. Some may be much less desirable than others. Still, if enough runs are undertaken, a subset of superior solutions can be created. Furthermore, the results from all runs may be added together to discern general trends in the selection process.

5.2.3 Terrestrial and Freshwater Spatial Analysis

5.2.3.1 SITES Parameters

Several factors besides the number and type of targets used influence Sites outcomes. These include type of planning units, protection status of planning units, planning unit cost measure, penalty applied for failure to meet target goals (“species penalty factor”), penalty applied for dispersed rather than clustered planning units in results (“boundary length modifier”), the number of repeat runs of the algorithm (and number of iterations within each run) to include in summing results from several scenarios, and goal level for each target.



Page 125

5.2.3.2 Planning Units

We used 500-ha hexagons selected from a layer that covers all the CIT and CFM planning areas. This not only allows consistency with the marine analysis, but using uniform sized planning units also avoids the area-related bias that can occur in the SITES planning unit selection process when differently sized planning units, such as watersheds, are used. The team recognized that the smaller in area the planning unit, the more spatially explicit the outputs could be, however, this needed to be balanced against computational constraints and limitations in resolution of data inputs. At the scale of the study area, 500 ha created a total of 27,888 planning units and represented the near maximum threshold for the computing power employed by the analysis team.

5.2.3.2.1 Planning unit status

We used four different protected areas scenarios for the planning units: no protected areas “locked in” the outcome, existing protected areas locked in, existing plus candidate protected areas locked in, and existing plus candidate plus option areas locked in. Candidate and option areas were only available for the Central Coast region.

5.2.3.2.2 Suitability index

Planning units with lower levels of human impacts should be chosen over those with higher levels of impacts, when other factors are equal. This general rule should lead to selection of areas that are more likely to contain viable examples of species and ecological systems. Thus, rather than simply using the number of hectares in each planning unit for the Area component of our SITES analyses, we developed a suitability index (i.e., cost index based on the same human impact data used in identification of intact landscapes.

Human impacted area was calculated by combining human-altered area (clearcut, urban, agriculture) with a 200 m buffer area around roads. Overlapping areas were treated as impacted (i.e., overlaps were only counted once), which also allow calculation of overall percentage of development (e.g., Moore 1991) in any planning unit or watershed.

To account for planning units with relatively little vegetated productive areas (and consequently little developable area and little productive habitat) we used the following suitability index:

Cost Index = Planning Unit Area + Planning Unit Area * Human Impacted Area/Potential Vegetated Area

with all areas measured in hectares. Potential Vegetated Area was calculated as the sum of vegetated habitat plus the sum of clearcut and urban areas. This assumes that existing development took place on formerly vegetated habitat areas. Note that this calculation omits bare rock, glacier and lake areas.

With this index applied, planning units with no human impacted area were given a cost of 500 ha, while those having all potential vegetated area impacted had a cost of 1000 ha and partially impacted planning units had cost between 500 and 1000 ha. Because the Sites algorithm seeks to minimize total portfolio cost, it selects planning units with low cost unless higher cost planning units contain targets that cannot be found elsewhere.



Page 126

5.2.3.2.3 Species penalty factor

Because we had no way to weight targets differently, we used the same penalty factor (one) for all targets.

5.2.3.2.4 Boundary length modifiers

We used boundary length modifiers (BLM) of 0.001, 0.01, and 0.1 to include a range of planning unit clustering in our final combined sum runs. Our boundary length modifier sensitivity analysis indicated that with a BLM of 0.001 there would be approximately 700 clusters of hexagons; with BLM 0.01, approximately 500 clusters; and with BLM 0.1 approximately 200 clusters. This range of BLMs allows us to explore issues of several small clusters versus fewer large clusters.

5.2.3.2.5 Repeat runs

We made 20 repeat runs (each comprised of 1,000,000 iterations of planning unit selection) for each of 15 combinations of boundary length modifier (three levels) and goal (five levels) for each of the four protected area scenarios. Thus, for each protection scenario we used a sum of 300 sites runs that resulted from 300,000,000 iterations of the simulated annealing algorithm. Hexagons chosen frequently represent places more necessary for biodiversity conservation, while those chosen few times represent locations where similar biodiversity is found many other places or where human impacts are significant.

5.2.3.2.6 Goal settings

We used five different SITES goal settings: 30%, 40%, 50%, 60%, and 70%, as described in Section 3.2. For terrestrial systems, these goals were applied against total available area of ecosystem types while freshwater goals were applied as percentages of total stream length. In the case of focal species, variability in habitat suitability and capability was addressed by applying the percentage goals to the total sum of habitat values for each model grid cell, for each species within the study area.

5.2.4 Nearshore Marine Spatial Analysis

5.2.4.1 MARXAN Parameters

For the analysis we used a site selection algorithm called MARXAN, written and developed by Ian Ball and Hugh Possingham. We set specific MARXAN parameters to increase efficiencies and optimization during site selection (see Pressey et al. 1996; Possingham et al. 2000; Leslie et al. 2002 for more information on the marine reserve selection algorithms). Shoreline ecosystems were analyzed within their native spatial format, linear segments determined by landform and slope. These units vary widely in length and extend over the entire coastline, providing a “natural” unit for analysis (Longley et al. 2001) as opposed to a unit of fixed length. Forage fish spawning sites and seabird colony data were attributed to the shoreline segments to represent the nearshore zone.



Page 127

We developed a framework for nearshore marine analysis using this linear spatial format. This analysis is coupled with expert review from the initial technical workshop in April 2003, and will be used for subsequent expert review workshops.

5.2.4.1.1 Planning units

We initially considered incorporating all the nearshore information in to hexagon planning units, but we found that MARXAN favored those hexagons with more shoreline and not necessarily “more efficient” shoreline toward meeting conservation goals. Hexagons arbitrarily fragment and aggregate shoreline units, leaving some hexagons with slivers of shoreline and others with large amounts (i.e., bays, inlets). In addition, the hexagon size lacks ecological justification and straddles narrow water bodies, therefore further aggregating shoreline units associated with different landmasses. Shoreline planning units tend to be more spatially explicit, easier for experts to review, and more intuitive in designing a conservation portfolio site.

5.2.4.1.2 Suitability index

To select priority conservation areas we included in the analysis a suitability index, or “cost index” as its referred to in MARXAN, which tends to reduce representation in places where human uses or modifications restrict conservation and modified options. The index was developed to calculate both the total cost of choosing a network of sites, and the relative cost of each planning unit.

Costs are usually referred to as impacts to the environment, making particular places less suitable for conservation. There are also jurisdictional costs where assumptions are made for different lands and waters holding a specific political status. These jurisdictional costs can be seen as more (i.e., lands already in conservation status) or less (i.e., lands devoted to resource extraction) suitable.

The index developed for the nearshore analysis consists of impact cost parameters only. We used the human impacted area information developed by Round River (see Section 6 for a compete discussion of the terrestrial cost parameters), which combined human altered areas (clearcuts, urban areas, agricultural lands) with a 200 m buffer around roads. We added three other cost parameters to this impact calculation: aquaculture tenures, enhancement facilities, and hatcheries.

The aquaculture tenures were spatially delineated as polygons, and were incorporated into the human altered areas. For enhancements and hatcheries, we buffered the point data by 100 m (the approximate minimum mapping unit diameter to convert points to polygons at a 1:250,000 scale). These were also included in the altered areas. All human impacted areas were given a score of 1; aquaculture tenures, enhancements, and hatcheries were given a 2 (Map 23).

Next we had to develop a human impact score for all shoreline planning units. To do this we performed raster-based analyses. A raster is a rectangular array of equally spaced cells, which taken as a whole represent thematic, spectral, or picture data (Zeiler 1999). All impact data was transformed into raster data, allowing us to calculate the human impact score for each planning unit. We first chose a focal mean function to assign a cost value to all shoreline cells. Focal functions compute an output cell value from those input cells that are within a “neighborhood”



Page 128

centered on the output cell. The neighborhood is defined as all cells within a given radius; all shoreline cells calculate a mean value from the surrounding upland and nearshore neighborhood. Once each shoreline cell has a mean value, we summarized all values per shoreline planning unit. This mean value across all shoreline cells gave us the human impact score from 0 to 2 (Map 23a).

To build the cost index, we simply took the mean planning unit length and the sum length of human impacts within that planning unit:

Cost = Mean Planning Unit Length + (Mean Planning Unit Length * Human Impact Score)

The mean planning unit length was 450 m. This represented the base value for all planning units; the maximum cost value was 1,190. We decided to take the mean value of all shoreline planning units instead of simply calculating cost equal to length. There is an assumption that larger areas have more impacts to them and therefore should have higher cost values. We do not believe this is true in the marine realm, where long stretches of coastline can be relatively pristine, while shorter lengths may be associated with high impacts. Each ecologically based planning unit that does not have equal areas (i.e., watersheds and shorelines) may need to be treated differently. For shorelines, giving all planning units equal value as a base assumes they have equal impacts to them; alone by multiplying the human impact score with the mean length do some planning units get a higher cost.

5.2.4.1.3 Boundary length modifiers

A boundary modifier determines the amount of clumping between individual planning units. This is usually done with polygons, but we customized the data inputs here to use the linear spatial format.

We developed a linear boundary modifier that clumped, or attached adjacent linear segments (arcs) along the shoreline. The algorithm was therefore able to assemble small fragments of shoreline into more continuous stretches (i.e., select an entire islands’ shoreline). We also felt that this helped us visualize more distinct sites across the shoreline ecosystem representation. We set the boundary length modifier to 0.1 for all MARXAN scenarios.

5.2.4.1.4 Species penalty factors

Setting the species penalty factor, or the conservation penalty factor (CFPF) as it is referred to in MARXAN, determines the priority with which the algorithm meets an individual target goal. The CFPF is a multiplicative factor, which can be unique to each conservation feature. It is primarily based on the relative worth of that conservation feature but it includes a measure of how important it is to get it fully represented. We set this factor based on the importance of the target; a high penalty factor is assigned to a high priority target to make the algorithm choose sites in potentially “high cost” areas to meet its goals.

All coarse filter shoreline ecosystem targets had a CFPF of 2. All coarse filter intertidal vegetation and community targets also had a factor of 2; fine filter seabird colony targets had a factor of 3. Man-made and undefined shoreline categories contained a factor of 1, with goals set to 0 to ensure the algorithm did not select them.



Page 129

5.2.4.2 Nearshore Marine Portfolio Assembly

The purpose of our efforts was to develop a conservation portfolio that, if conserved and properly managed, will protect a representative subset of the existing nearshore marine biodiversity in the coastal zone of the Coast Information Team study area. This representative subset encompasses the intertidal zone and shallow subtidal zone.

We made the choice early on in the planning process to run separate analyses for the offshore, nearshore and terrestrial environments. This limited our ability to analytically test the relationship between nearshore and terrestrial ecosystems, and therefore have to rely on experts to review the separate outputs in delineating and integrating sites across ecosystem types.

The approach for building a nearshore marine portfolio combined expert input with spatial analysis. This systematic approach of using expert input at the beginning, and throughout, the planning process was used to test preliminary analytical results while refining the portfolio.

5.2.4.3 Initial Seascape Sites

One of the objectives from the technical workshop was to select biologically diverse coastal sites based on expert knowledge. These initial seascape sites were chosen to capture relatively large, intact ecosystems that represent the region’s nearshore biodiversity. The experts were asked to select areas along the B.C. coast using the following three criteria:

1. Large nearshore sites are important for marine biodiversity

2. Sites must be stratified12 across the ecoregion

3. There should be diversity in the types of sites chosen, i.e., good for birds, good for invertebrates, etc.

The results were the identification of 18 seascape sites:

1. Mouth of Skeena River

2. Mouth of Nass River

3. Dundas Island

4. Langara Island

5. Forrester Island

6. Bowie Shoal

7. Dogfish Bank

8. Naikoon Park

9. Gwaii Haanas

12 Stratification is a process of creating subsections within the ecoregion for the purpose of spreading the site selection and review process across the entire water body. The delineation of subsections also serves as a surrogate for oceanographic processes including salinity, current, and temperature.



Page 130

10. Clao Bay

11. Princess Royal Island

12. Goose Group

13. Hakai Pass

14. North Queen Charlotte Strait

15. Scott Island

16. Broughton Archipelago

17. Johnstone Strait

18. Redonda Island/Oyster Inlet

The identification of these seascapes helped to guide subsequent analysis. The intention is that these areas will be tested and refined using more spatially explicit data and the analysis described here.

5.2.4.4 Spatial Analysis

The purpose of the spatial analysis was to implement and evaluate 12 different MARXAN scenarios to test the irreplaceability and sensitivity of the site selection.

Irreplaceability analyses indicate which sites are consistently chosen. Planning units that get chosen the most often are the least replaceable. These analyses can be evaluated through the “summed solution” output, which is the result of the number of times an individual planning unit gets chosen within a scenario. The output from all 12 scenarios was combined to form the sum of summed solutions, or the summed solution gradient. With each of the 12 scenarios run 20 times, we had a gradient from 0 to 240 solutions. Within each scenario the algorithm did 1,000,000 iterative selections per run.

Two multiple goal scenarios were run: one optimizing for 10% of the entire shoreline, and another optimizing for 20%. When shoreline ecosystem targets were set to 10%, intertidal vegetation, communities, and fine filter targets were set to 15%. When shoreline ecosystem targets were set to 20%, intertidal vegetation, communities, and fine filter targets were set to 30%. These were applied to all stratification schemes and cost/no cost scenarios. We used three stratification schemes to divide the nearshore zone: marine ecosections, project regions, and no stratification. Marine ecosections provided a stratification scheme based on oceanic and coastal processes. Project regions divided the nearshore by surveying and data collection efforts; and a third scheme was to run the algorithm with no stratification. We also ran the algorithm with and without the cost index (without the cost index means that all planning units were given a cost of 1).

The 12 scenarios are described as follows:

1. 10% shoreline, marine ecosections, cost

2. 10% shoreline, marine ecosections, no cost



Page 131



5. 10% shoreline, project regions, cost

6. 10% shoreline, project regions, no cost



9. 10% shoreline, no stratification, cost

10. 10% shoreline, no stratification, no cost



Sensitivity analyses are conducted by selecting thresholds in the summed solution gradient and evaluating how much representation of the nearshore targets are captured. The area required to meet goals changes as we lower the threshold along the summed solution gradient. Sometimes the area required changes very little as we lower the threshold; sometimes it increases drastically. In evaluating sensitivity, it helps to know the shape of this curve. The summed solution gradient from the 12 MARXAN scenarios provided the base for performing the sensitivity analysis.

There are several ways we are reporting on the results of the analysis:

• Display the summed solution gradient as the raw output of the nearshore analysis. Compare this gradient with results from the technical workshop.

• Choose the most optimized “best” run from the 12 scenarios. This will be used to represent the full array of coarse and fine filter targets and an overarching shoreline length captured. Compare this run with results from the technical workshop.

• Find thresholds along the gradient that capture the multiple goal levels set for shoreline ecosystem, intertidal vegetation and communities, and fine filter targets.

• What representation do we capture when we set a threshold of the upper 10% of the entire shoreline length? This equates to the “High Conservation Risk Option,” or the high priority nearshore sites.

• What representation do we capture when we set a threshold of the upper 20% of the entire shoreline length? This equates to the “Medium Conservation Risk Option,” or the medium priority nearshore sites.

• What representation do we capture when we set a threshold of the upper 30% of the entire shoreline length? This equates to the “Low Conservation Risk Option,” or the low priority nearshore sites.

• Take the established thresholds in item 3 and buffer the selected shoreline segments to create a “nearshore reporting-out unit.” This adds more spatial explicit prioritization within the high, medium and low conservation risk options. Buffering selected shoreline segments by



Page 132

1 km (the relative distance from the mean high water line that encompasses the nearshore zone down to 20 m depth) incorporates non-selected segments.

• Calculate the total length of selected shoreline in each nearshore reporting-out unit. This gives priority to the longer segments.

• Calculate the density of selected shoreline by the total length within each nearshore reporting-out unit. This gives more equal priority to shorter and longer segments.

• Calculate the mean weighted average by scoring each shoreline planning unit and multiplying it by the total length in the nearshore reporting-out unit. This lowers the priority of longer segments, and heightens priority to shorter ones.

5.2.4.5 Expert Review

We have not formally presented the preliminary results of the analysis for expert review. We hope to do this in the coming months. We plan to reconvene experts from the technical workshop and others to review and verify preliminary results of the analysis, and make suggestions for next steps. However, we have output from the spatial analysis to compare with the 18 initial seascape sites. That comparison is described in the next section, “Results.”

5.2.5 Marine Deep-water Spatial Analysis

5.2.5.1 Marxan Parameters

Marxan consists of 8 main variables to direct the optimization algorithm:

1. Conservation Targets (Goals): How much of a feature is aimed for in the MPA network.

2. Penalty Values: How much cost is accrued for not attaining the conservation target.

3. Boundary Length Modifier: The relative cost of a reserve’s perimeter

4. Minimum Separation Distance: The minimum distance that distinct groupings of a feature should be from one another.

5. Separation Number: The number of distinct groupings of a feature required (i.e., replication).

6. Minimum Clump Size: The minimum number of planning units (hexagons) needed to count as a valid grouping of the feature.

7. Planning Unit Cost: A relative value applied to planning units such that some may be more difficult or “expensive” to set aside than others.

8. Boundary Cost: The relative cost of the planning units’ shared boarders.

Of these, the first three are the most important. The first parameter, conservation goals, has been discussed above (Section 3), and is equivalent to stating how much of a feature is enough to meet one’s conservation objectives. In the marine ESA, we explored a wide variety of goals so as to provide planning tables with a range of possibilities, from low to high conservation objectives.



Page 133

5.2.5.2 Planning Units

The Marine ESA planning units are a regular grid of 500-ha hexagons. There are about 32,000 of these hexagons in the analysis, which covered the entire CIT marine study area, and down the west coast of Vancouver Island.

To get an accurate picture of how abundant a feature is within a planning unit (hexagon) we considered two factors:

1. How much of it is there

2. How much of it could there be there (i.e., its possible maximum). In our analysis this often equals the amount of seawater contained in the hexagon, but for shoreline features would be a total measure of shoreline per hexagon.

Considering just the summation of a feature’s presence (point #1) would unfairly penalize hexagons that had full 100% presence of the feature, but not 100% water. This situation might prove to be important when, for example, the nearshore component plays a critical role, such as in estuaries. In this situation, a planning unit is very unlikely to contain but a fraction of its area as water, and yet may play a far more important functional role than an offshore planning unit with the same amount of the feature, but surrounded by water.

In our model we make allowances for how much water is available per planning unit, accounting for feature density, as well as occurrence.

Presence/Absence Area Data

For presence/absence data, the formula we generally used is:

HexScore f (presence) = √((∑f) 2/(2 Nf)) ……1

Where f is the feature occurrence (presence = 1, absence = 0); thus ∑f is the sum of all feature cells;

And Nf is the total number of possible feature cells −, which is usually the same as the total number of water cells.

Another way to state this is:

HexScore f (presence) = √(Σf * fmean)/2 ……2

Where fmean is the mean value of that feature, wherever there is water. For presence data, this is the same as density as discussed above.

For presence/absence features, the scores can range from 0 to 16 per hexagon. This compression of values was found to be robust to random grid shifts and variations in base shorelines used by different data sets.

For weighted (“Relative Importance” −RI−) features, the above formula is multiplied by the mean of the feature cell weightings:

HexScore f (RI) = HexScoref (presence) * RImean ……3

Where, RImean = ∑f (RI)/Nf (presence);



Page 134

And ∑f (RI) is the sum of all the RI feature cells

And Nf (presence) is the total number of presence feature cells.

5.2.5.2.1 Line and point features

The above formulae were used for most of our two-dimensional area features (GIS “polygons”). For line features, we used the same formulae, except that Nf represents the total number of possible shoreline cells, instead of water.

Point features were all given buffers to convert them into appropriate areas, and then were treated as above.

5.2.5.3 Penalty Values

Assigning a penalty to a feature is in effect saying how much it matters if this feature’s goal (target) is not met. That is, for features that do not meet their goals, penalties are assigned (on a sliding scale based on how closely the goal was achieved); and in turn it is these penalties that will “direct” the algorithm in its search for features. Thus, features with higher penalties are generally met first (if they can be met) than similarly distributed features with lower penalties. Generally, we used the penalty value as a relative factor to reflect the relative importance of a feature, and sometimes to also reflect the relative confidence in that data set or its spatial completeness, as compared with others. We assigned lower penalties to those data sets in which we had lower confidence. We did not want these data sets driving the analysis. We assigned higher penalties to rare, threatened, and endangered species, as well as to features that play important ecological roles (such herring spawn).

As with goals (targets), penalties were first given a relative ranking. From those, weightings were assigned as follows:

Relative penalty Marxan weighting

Low 0.25

Mod-Low 0.50

Moderate 1.00

Mod-High 2.00

High 4.00

V. High 8.00

Appendix 3.1 lists all 93 features in the marine ESA, and their assigned relative penalties.

5.2.5.4 Boundary Length Modifier (Clumping)

Boundary Length Modifier (BLM): The relative cost of a reserve’s perimeter. Higher costs will force larger (but fewer) reserves, whereas a low cost will allow for several small ones. We have explored a wide range of this parameter (BLM= 0.004, 0.008, 0.016… 8.000) but have focused on



Page 135

four to cover the range from fragmented to moderately clumped (BLM= 0.0625, 0.250, 1.000, 4.000) This is an arbitrary parameter that must be arrived at through experimentation. While we found that solutions using a BLM near 1.0 offered good efficiency with realistic manageability, we also discovered that the more fragmented solutions (which more truly represented the densities of conservation values) were valuable when summed together to show trends or “hotspots.”

As solutions progressed from scattered to clumped, they behaved predictably, shedding smaller reserves and aggregating onto the larger ones. This would indicate that the data populate the planning units in a consistent fashion and that the planning units themselves are consistent.

5.2.5.5 Other Parameters

The other Marxan Parameters were handled as follows:

• Minimum Separation Distance: Not used. This parameter greatly increases processing requirements. For such a large number of planning units (32,000) and features (93), its use was impractical.

• Separation Number: Not used. (As above.)

• Minimum Clump Size: Not used. We felt the 500-ha hexagons were already sufficiently large. In practice, the hexagons naturally clump together.

• Planning Unit Cost: All planning units treated the same. Cost set to 1. As that the objective of this exercise was to explore just conservation values, we did not consider whether some planning units might in practice be more difficult to protect than others.

• Boundary Cost: This parameter was used to fine-tune the relative clumping of hexagons in the four Ecological Regions (inlets, passages, shelf, slope). To determine this value we looked at the edge to area ratio of each of these regions and then created an appropriate scalar. The non-dimensional measure we used was: √(P2/A where P = total perimeter of region, and A = total area of the region. By altering the boundary costs per region, we allowed for more fragmented solutions in areas constrained by geography, such as inlets, but encouraged more clumped solutions in open waters, such as over the continental slope. The resulting boundary costs were as follows:

Region Boundary cost

Continental Slope 1.54

Continental Shelf 1.00

Passages 0.34

Inlets 0.21



Page 136

5.4 Options and Scenarios

5.4.1 Background

The use of site selection algorithms such as SITES and MARXAN brings a degree of flexibility to the spatial analysis process. Multiple settings for goals, “locking in” selected areas, adjustments to “clumping,” and the use of a cost function that incorporates human impact information, are all tools that build choice into the analytical process. Decision-makers and stakeholders, as well as planners, can explore these choices, and the subsequent implications. To facilitate examination of spatially explicit conservation solutions, we made use of SITES and MARXAN to create a series of potential conservation solutions that in combination, are referred to as Options and Scenarios. At the heart of this exercise was an attempt to prioritize solution outputs based on criteria related to conservation utility, based on target information and SITES/MARXAN outputs, and conservation condition, based upon of the evaluation of human impacts within the study area.

5.4.2 Conservation Utility

A key concept in conservation planning is irreplaceability (Pressey et al. 1994; Margules and Pressey 2000; Pressey and Cowling 2001). Irreplaceability provides a quantitative measure of the relative contribution different areas make toward reaching conservation goals, thus helping planners choose among alternative sites in a portfolio. As noted by Pressey (1998), irreplaceability can be defined in two ways: 1) the likelihood that a particular area is needed to achieve an explicit conservation goal; or 2) the extent to which the options for achieving an explicit conservation goal are narrowed if an area is not conserved. Given the constraints under which the site selection algorithms operate, we can expect that summed solutions will describe a range of important conservation criteria including rarity, richness, diversity and complementarity. Optimization of these criteria is approximated through the selection of a minimum set of planning units to meet goals for our conservation targets. For the CIT ESA, we have used these summed solutions as a broad measure of irreplaceability, which for the purposes of this report, we more simply describe as “conservation utility.” Conservation utility however is not always a direct and absolute measure of true irreplaceability since areas with high conservation utility may indeed be replaced by using larger areas of lower utility.

In the case of terrestrial and freshwater analysis, a combination of 5 goal settings and 3 boundary length modifiers were repeated 20 times each for a total of 300 possible conservation solutions, each of which were integrated into a single final summed solutions. A conservation utility score was derived directly from the frequency by which any one planning unit was selected in these 300 repetitions, such that a unit selected in every solution received a score of 300, while a unit never selected was scored as a zero. For evaluation purposes, the data were then segregated into 5 classes based on an equal interval distribution of summed run selection scores of the summed solutions (see Table 5.1). A planning unit selected 180 or more times in SITES fell into the top 2 quintiles, or top 40%, of the solution and was scored as having high conservation utility. Analysis units with scores in the middle quintile were scored as medium conservation utility, and those in the 4th quintile were scored as low conservation utility. Those units with a score in the lowest quintile were not ranked.



Page 137

Table 5.1 CIT ESA conservation utility ratings based on SITES summed solution scores

Summed solution score Quintile Conservation

utility rating

∃ 180 Top 2/5 High

< 180, ∃ 120 3/5 Medium

< 120, ∃ 60 4/5 Low

< 60 5/5 Not ranked

5.4.3 Condition

Condition was used as a surrogate measure for target viability in the CIT ESA and was evaluated using the human impacts information described in Section 4. The six impact classes described in Table 4.2 were simplified into the three broad condition classes—intact, modified, and developed (see Table 5.2).

5.4.4 Alternative Options for Conservation

To explore the interaction between conservation utility and condition, analysis units were clustered into 3 conservation tiers based on the conservation utility and condition matrix in Table 5.2. Under this framework, areas ranked as intact or modified that also hold high conservation utility, or intact areas with medium conservation utility, were ranked as Tier 1. The middle tier (Tier 2) represents those areas with high utility but which are highly impacted, or areas with low utility, but which are intact, or areas that fall within the mid-range of both criteria (medium utility/modified condition class). Tier 3 represents those analysis units that are developed and which have a medium or low conservation utility.

Table 5.2 Conservation tiers for the CIT ESA analysis units

Condition

Intact Modified Developed

High 1 1 2

Medium 1 2 3

Con

serv

atio

n U

tility

Low 2 3 3



Page 138

5.4.5 Alternative Scenarios Analysis

The Central Coast LRMP table has already proposed several potential land use scenarios, and we wanted to evaluate each in terms of their performance against the three conservation options being generated by the ESA. To facilitate this comparison, the protected areas described by each scenario were locked into the SITES solution for terrestrial/freshwater runs. Four alternative land use scenarios were evaluated as follows:

1. Unconstrained Analysis

2. Base Case – existing protected areas locked into SITES.

3. Candidate Case – existing protected and CCLRMP Candidate Areas locked into SITES.

4. Option Areas Case – existing protected areas, CCLRMP Candidates, and Option Areas locked into SITES.

5.4.6 Comparing Options and Scenarios

To compare between and within options and scenarios, potential solution sets from summed runs were evaluated against three goal thresholds, 30, 50, and 70%. For each scenario, performance of conservation tiers (options) was assessed both in terms of effectiveness, as measured by the proportion of targets that met or exceeded the goal threshold, and efficiency, the proportion of the study area required to meet the threshold.

5.4.7 Relative Biodiversity Index (RBI)

After completion of the initial draft of the ESA, it became apparent that a number of supplementary materials were required to directly support ongoing negotiations at LRMP tables. In particular, stakeholders re-iterated the limitations with regard to making use of a single optimized conservation solution. Instead, political realities demanded a sequential building of a protected areas system based on the progression of a series negotiating steps. To support decision-making throughout these steps, stakeholders also requested more transparency in the presentation of data on conservation values. Specifically, LRMP table members were interested in maps that showed biodiversity values broken out by system type and species, and asked that these values be displayed on a continuum, using third order watersheds as a reporting unit.

To meet these needs, the ESA team created a second index of conservation value based on the existing data inputs that were used in SITES runs, and applied it to third order watersheds. This methodology involved calculating for each ESA target the relative contribution of each watershed to the total available habitat or ecosystem extent within each LRMP boundary. These calculations were based for each watershed on the following formula:

RBItarget = target system Ha (or habitat value) within watershed total target Ha (or habitat value) in LRMP

These scores were calculated independently for each focal species and for each ecosystem type. They were then also normalized out of a score of 1 and summarized into a total focal species score as well as a total freshwater systems score, salmon/steelhead score, old-growth ecosystems score, and total focal terrestrial ecosystems score. Finally, these were translated into a total relative biodiversity index (RBI) for all third order watersheds in each LRMP area.



Page 139

6.0 SPATIAL ANALYSIS: RESULTS AND DISCUSSION

6.1 Terrestrial/Freshwater Spatial Analysis Results

Importantly, site selection algorithms, by themselves, do not address the more difficult and real-world questions concerning the area needed to maintain viable populations of species and the persistence of biodiversity. The following results can however provide stakeholders and decision makers with reference points for exploring how potential conservation solutions might be distributed across the study area. This information can help direct where conservation values are highest and which geographies might best be suited for no or low risk landscape management practices. Conversely, this information may be useful in helping to identify locations where resource development might occur with minimal impacts to ecological values. The following sections provide a number of possible frames through which the combined mapping and analysis of the ecological value information described so far in this report might be viewed.

6.1.1 Options and Scenarios measured by Analysis Unit

Summed run solutions for each of the four land use scenarios are displayed in Maps 32 through 35 and priority conservation tiers are shown in Maps 36 through 39. While the pattern of conservation utility differs between each scenario, there are only small differences between the overall performances of the solutions. Performance measures included both the solution’s efficiency, measured by the amount of area swept into each conservation option or tier, and effectiveness, measured by the conservation goals reached by those tiers (see Figure 6.1 and Table 6.1).

Table 6.1 Comparison among option and scenario combinations for CIT ESA analysis units

Proportion of targets meeting identified goal threshold Option*/Scenario** 30% 50% 70%

% of study area in solution

% of solution protected

1.A 0.77 0.56 0.20 40.1 13.6¹

1.B 0.76 0.53 0.18 42.0 26.3

1.C 0.77 0.52 0.18 41.7 33.9

1.D 0.76 0.52 0.18 41.9 43.0

2.A 0.99 0.91 0.70 61.9 12.5¹

2.B 0.99 0.92 0.67 61.9 17.8

2.C 0.99 0.92 0.66 61.3 23.0

2.D 1.0 0.92 0.65 60.8 29.6

3.A 1.0 0.99 0.88 68.3 11.5¹

3.B 1.0 0.99 0.87 68.1 16.2

3.C 1.0 0.98 0.96 67.4 21.0

3.D 1.0 0.98 0.86 66.6 27.0

¹Compared with existing protected areas.



Page 140

*Options:

1. Tier 1 analysis units 2. Tier 1 and 2 analysis units 3. Tier 1,2,3 analysis units **Scenarios:

A. Unconstrained analysis B. Base Case – existing protected areas locked in. C. Candidate Case – existing PAs and candidate areas locked in D. Option Areas Case – PAs, candidates, and option areas locked in

Figure 6.1 Conservation tiers for CIT ESA analysis units, compared between protected area scenarios.

0

10

20

30

40

50

60

70

% o

f stu

dy a

rea

No areas locked in Existing protectedareas locked in

Exisiting protectedareas and candidate

areas locked in

Existing protected,candidate, and option

areas locked in

Protected Area Scenario

TierThreeTier Two

Tier One

Protected

Depending on the conservation scenario, between 44 and 50% of the highest conservation utility analysis units (equal to 44–50% of the study area) were required to satisfy the 30% representation goal threshold for most targets. As seen in Figure 6.2, increases in solution area bring about proportional increases in solution effectiveness. To achieve 50% goals for most conservation targets, as much as 60–70% of the study area was required, with inefficiencies emerging at the top end of a sigmoidal curve, signifying that after approximately 50–60% of the study area (50–60% of the highest utility analysis units) has been incorporated into the solution, subsequent improvements in meeting goals require proportionally larger areas of land and water. This effect is much more pronounced when the 70% representation goal threshold is examined. In this case, inefficiencies arise to the point where the inclusion of more planning units yields only minute improvements relative to meeting goals.



Page 141

Figure 6.2 Progress toward representation goals for CIT ESA SITES summed solutions: Existing protected areas locked in.

0

0.2

0.4

0.6

0.8

1

1.2

10 20 30 40 50 60 70 80 90

Percent of study area in conservation solution

Prop

ortio

n of

targ

ets

for w

hich

repr

esen

tatio

n go

al is

reac

hed

or e

xcee

ded

705030

Representation Goal

Note that using the SITES summed solution scores for identifying core conservation areas, does not represent an “optimal” design methodology. So called optimal SITES designs are created using the “best run” feature of SITES, but given the considerable uncertainty around conservation goals, boundary length modifiers, and the direction of LRMP negotiations, the team elected to use a summed run approach so as to convey a range of potential conservation solutions as opposed to a single “correct” solution. Providing an optimal or best solution would be appropriate if planning tables were prepared to accept a single conservation solution in its entirety, however, from the beginning of the planning process, such an outcome was obviously unrealistic given the competing interests among stakeholders. It is recommended that as agreements around protected areas are sequentially reached, that SITES optimal runs be used as a means of evaluating the success of proposed reserves in meeting representation benchmarks, and for identifying remaining areas needed to for conservation. If protected areas are chosen based on high conservation values, these optimized runs should yield more efficient solutions i.e., the percentage of area captured in the SITES run should be closer to the percentage of total ESA targets meeting representation goals.

6.1.2 Protected Areas: Existing and Potential

The current and potential conservation contribution of each land use scenario is presented in Figure 6.3. At a representation goal of 30%, we see that less than 20% of the conservation elements being targeted by the CIT ESA meet the goal threshold—under any combination of proposed and existing protected areas. However, it is apparent that there are considerable



Page 142

conservation values within the areas that have been designated and proposed. Table 6.2 shows that up to a quarter of the regions Tier 1 analysis units would be captured in a scenario that designated candidate and option areas as protected.

Figure 6.3 Proportion of CIT ESA Conservation targets represented in existing and proposed protected area scenarios based on 5 example representation thresholds.

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

12% 30% 50% 70% 100%

Representation threshold

Prop

ortio

n of

targ

ets

mee

ting

repr

esen

tatio

n th

resh

old

Existing ProtectedAreas

Existing plusCandidates

Existing plusCandidates andOption Areas

Protected Area Scenario

Table 6.2 Protected area status of analysis units from the “No areas locked in” SITES solution.

Conservation Tier

% of study area in solution

% of solution in existing protected areas

% of solution in existing protected areas plus Candidate Areas

% of solution in existing protected areas plus Candidate and Option Areas

1 40.9 13.6 19.0 25.0

2 61.9 12.5 16.9 22.2

3 68.3 11.5 15.5 20.4

6.1.3 Relative Biodiversity Index

Results of total RBI scores for third order watersheds are displayed in Map 40. High total RBI scores may occur as a result of many co-occurring conservation targets in a watershed, but may also result when a watershed contains most of the extent (or available habitat value) for a single



Page 143

target. This latter instance occurs when a watershed contains examples of targets that are more spatially limited within the study area. This methodology also favours larger watersheds over smaller since it is a cumulative score and the more area available, the higher the likelihood that more ecosystems and habitat will be present.

These RBI scores improve the understanding of the distribution of conservation values within the LRMP table’s specific sphere of influence (i.e., the specific LRMP boundary) and present the data in a format that is more intuitive and accessible (i.e., watersheds as opposed to hexagons). These scores are also useful for interpreting ESA results where questions exist as to why any particular area of the region is rated as high conservation utility. RBI scores allow the user to determine which of the ESA’s conservation targets is influencing site selection—freshwater or terrestrial, ecosystem or focal species, and further, which specific systems or species are driving selection. Detailed information for systems and individual species RBI is available to stakeholders through CIT ESA data sets.

Since condition of the watershed is not incorporated into the measure as displayed in Map 40, some high conservation value watersheds classified as “developed” score higher than expected when compared to conservation priorities emerging from SITES runs. In the future, the authors recommend that the RBI score be adjusted by the watershed’s condition class to account for impacts.

6.1.4 Representation Analysis of CC LRMP Negotiated Agreement

One of the most simple but effective uses of ESA information should be to assess emerging conservation proposals based on how well proposed protected areas represent conservation targets—a process often referred to as “gap” analysis. The ESA team had the opportunity to evaluate a number of proposed protected areas solutions emerging from the CC LRMP table in the fall of 2003. Our simple gap analysis was based on a calculation of how much of each system type and focal species habitat value was captured in each proposed solution (including previously established protected areas and newly negotiated areas) as compared to the total system type or habitat value available for the entire LRMP area.

Results from the analysis of the CC LRMP final agreement of December 12, 2003 are presented in Appendix 5.1 and a specific example, the analysis of focal species habitat representation, is presented in Figure 6.4. Previous to the December 12 agreement, existing protected areas in the region had represented 10% or less of ESA focal species habitat. However, the combination of candidate areas and option areas agreed to through the LRMP process increased representation of habitat by roughly three to five-fold, depending on the species. Substantial and important gains were also achieved for focal ecosystems, old-growth ecosystems, and freshwater systems (Appendix 5.2). A closer examination of focal species habitat differentiated by ecosection does however demonstrate significant shortfalls in representation of some ecosections of the CC LRMP area—most notably in the Outer Fjordlands ecosection. Similar shortcoming exist for focal ecosystem targets and old growth in this ecosection, and to a lesser extent, in the Northern Pacific Ranges ecosection, Nonetheless, the authors are unaware of other North American examples where a protected areas system has so specifically captured such a significant proportion of a region’s biodiversity values.



Page 144

Figure 6.4 Focal Species habitat representation based on CC LRMP agreement.

0

10

20

30

40

50

60

70

80

90

100

Black Bear Grizzly Bear Tailed Frog Mountain Goat MarbledMurrelet

NorthernGoshawk

Focal Species

% H

abita

t Rep

rese

ntat

ion

EXISTING PA CANDIDATE PA OPTION AREA

Information emerging from gap analysis for all three LRMP areas needs to be highlighted for future work around ecosystem-based management. Examples of habitats and systems not well represented need to be identified and well protected through restrictive management guidelines. While over half of ESA targets in the CC LRMP are represented to at least a 30% benchmark, many remain underrepresented against this measure. Further, the 30% threshold still lies well below the CIT’s Ecosystem-based Management Framework prescribed minimum of 70% representation of the range of natural variation of ecosystems. The 70% goal is the minimum threshold needed to meet the agreed upon “low-risk” management benchmark for the region. Presumably, mangers intend to make up this shortfall through the management of the ‘matrix’ between protected areas.

Given the gap between what is represented in proposed “no-risk” protected areas, and the over-all “low-risk” regional goal, the authors presume that management of the matrix between protected areas will need to be conservative to meet agreed upon regional objectives. It is also important to note that even those species and systems well represented within watersheds designated for protection may be inadequately conserved if protected areas are not well designed and managed. For example, continuation of commercial trophy hunting and mineral exploration/extraction might seriously impair the ability of these areas to maintain species and system viability. Next steps should include analysis of connectivity between proposed protected areas, and an examination of their size, configuration and management. Such work is critical for exploring questions related to the adequacy of proposed protected areas for ensuring the long-term viability of the species and systems represented within.



Page 145

6.2 Marine Nearshore Spatial Analysis Results

The purpose of the spatial analysis was to implement and evaluate 12 different MARXAN scenarios to test the irreplaceability and sensitivity of the site selection.

Irreplaceability analyses indicate which sites are consistently chosen. Planning units that get chosen the most often are the least replaceable. These analyses can be evaluated through the “summed solution” output, which is the result of the number of times an individual planning unit gets chosen within a scenario. The output from all 12 scenarios was combined to form the sum of summed solutions, or the summed solution gradient. With each of the 12 scenarios run 20 times, we had a gradient from 0 to 240 solutions. Within each scenario the algorithm did 1 million iterative selections per run.

Two multiple goal scenarios were run: one optimizing for 10% of the entire shoreline, and another optimizing for 20%. When shoreline ecosystem targets were set to 10%, intertidal vegetation, communities, and fine filter targets were set to 15%. When shoreline ecosystem targets were set to 20%, intertidal vegetation, communities, and fine filter targets were set to 30%. These were applied to all stratification schemes and cost/no cost scenarios. We used three stratification schemes to divide the nearshore zone: marine ecosections, project regions, and no stratification. Marine ecosections provided a stratification scheme based on oceanic and coastal processes. Project regions divided the nearshore by surveying and data collection efforts; and a third scheme was to run the algorithm with no stratification. We also ran the algorithm with and without the cost index (without the cost index means that all planning units were given a cost of 1).

The 12 scenarios are described as follows:













Sensitivity analyses were conducted by selecting thresholds in the summed solution gradient and evaluating how much representation of the nearshore targets were captured. The area required to meet goals changed as we lowered the threshold along the summed solution gradient. We set



Page 146

three thresholds to illustrate this: 10% (Option A), 20% (Option B), and 30% (Option C) of the entire shoreline length. The summed solution gradient from the 12 MARXAN scenarios provided the basis for setting these thresholds.

There are several ways we are reporting on the results of the sensitivity analysis:

• Display the summed solution gradient as the raw output of the nearshore analysis (Map 41).

• Display options A, B, and C by setting thresholds to capture 10%, 20%, and 30% of the entire shoreline length. These thresholds along the gradient captured the multiple goal levels set for shoreline ecosystem, intertidal vegetation and habitats, and fine filter targets.

• Option A - Nearshore: Upper 10% of the entire shoreline length. Those planning units with summed solution values ranging from 122 to 240 were included in this category (Map 42).

• Option B - Nearshore: Upper 20% of the entire shoreline length. Those planning units with summed solution values ranging from 74 to 240 were included in this category (Map 43).

• Option C - Nearshore: Upper 30% of the entire shoreline length. Those planning units with summed solution values ranging from 42 to 240 were included in this category (Map 44).

Target amounts captured in Options A, B, and C for the nearshore are detailed in Tables 6.3 and 6.4.



Page 147

Table 6.3 Coarse filter targets captured by nearshore Options A, B, and C

Case Target Target

_id Shore units

Total length (m) Goal

Goal amount

High 10% count

High 10%

length10% captured

Medium 20% count

Medium 20% length

20% captured

Low 30% count

Low 30% length

30% captured

1 Channel 1001 3 1,548.5 20% 309.7 0 0.0 0.0% 2 1,329.7 85.9% 3 1,543.0 99.6%

2 ChannelE 1002 9 2,690.1 20% 538.0 1 556.3 20.7% 1 556.3 20.7% 1 556.3 20.7%

3 ChannelP 1003 73 38,751.9 20% 7,750.4 2 3,282.0 8.5% 7 9,657.1 24.9% 9 11,402.6 29.4%

4 Estuary Wetland 1004 40 43,949.0 20% 8,789.8 3 11,193.

1 25.5% 5 16,070.9 36.6% 8 23,194.9 52.8%

5 Estuary WetlandE 1005 13 9,942.3 20% 1,988.5 1 207.1 2.1% 4 5,944.5 59.8% 6 8,215.1 82.6%

6 Estuary WetlandP 1006 1134 1,046,811.2 20% 209,362.2 21

94,940.7 9.1% 62 235,475.6 22.5% 115 353,632.9 33.8%

7 Estuary WetlandVE 1007 9 9,323.4 20% 1,864.7 1 428.6 4.6% 2 2,508.3 26.9% 4 5,016.9 53.8%

8 Estuary WetlandVP 1008 47 46,032.7 20% 9,206.5 2

12,229.2 26.6% 3 14,167.6 30.8% 5 16,784.1 36.5%

9 Gravel Beach 1009 215 78,725.4 20% 15,745.1 4 8,850.9 11.2% 11 14,529.0 18.5% 18 22,749.0 28.9%

10 Gravel BeachE 1010 333 93,229.1 20% 18,645.8 12 12,446.

9 13.4% 24 19,454.4 20.9% 40 28,924.7 31.0%

11 Gravel BeachP 1011 2983 941,273.4 20% 188,254.7 77 96,079.

1 10.2% 162 170,895.3 18.2% 279 254,010.4 27.0%

12 Gravel BeachVP 1012 12 4,336.2 20% 867.2 1 1,081.1 24.9% 1 1,081.1 24.9% 3 2,022.2 46.6%

13 Gravel Flats 1013 28 15,052.7 20% 3,010.5 1 1,580.3 10.5% 3 3,119.3 20.7% 4 4,528.8 30.1%

14 Gravel FlatsE 1014 56 19,683.4 20% 3,936.7 3 3,767.0 19.1% 6 5,732.1 29.1% 10 8,045.7 40.9%

15 Gravel FlatsP 1015 148 76,247.2 20% 15,249.4 4 5,582.7 7.3% 13 18,017.7 23.6% 20 24,584.8 32.2%

16 Gravel FlatsVE 1016 9 3,599.3 20% 719.9 1 2,115.1 58.8% 1 2,115.1 58.8% 2 2,491.0 69.2%

17 High Tide Lagoon 1017 50 85,740.2 20% 17,148.0 1

18,122.6 21.1% 3 25,669.6 29.9% 5 39,864.9 46.5%

18 High Tide LagoonE 1018 10 4,254.7 20% 850.9 0 0.0 0.0% 3 1,776.0 41.7% 3 1,776.0 41.7%

19 High Tide 1019 210 236,286.4 20% 47,257.3 3 14,308. 6.1% 11 77,215.4 32.7% 20 96,726.6 40.9%



Page 148

Case Target Target

_id Shore units


Goal amount

High 10% count

High 10%

length10% captured

Medium 20% count

Medium 20% length

20% captured

Low 30% count

Low 30% length

30% captured

LagoonP 7

20 High Tide LagoonVP 1020 60 80,393.8 20% 16,078.8 1

14,553.0 18.1% 5 49,823.6 62.0% 7 56,989.1 70.9%

21 Man-made (not a target) 1021 13 5,253.3 0% 0.0 0 0.0 0.0% 0 0.0 0.0% 0 0.0 0.0%

22 Man-madeE (not a target) 1022 5 1,262.8 0% 0.0 0 0.0 0.0% 0 0.0 0.0% 0 0.0 0.0%

23 Man-madeP (not a target) 1023 302 135,230.8 0% 0.0 0 0.0 0.0% 3 18,725.8 13.8% 5 25,781.5 19.1%

24 Mud Flat 1024 7 4,985.3 20% 997.1 2 2,468.2 49.5% 4 3,471.2 69.6% 5 3,788.2 76.0%

25 Mud FlatP 1025 156 107,010.0 20% 21,402.0 6 11,516.

1 10.8% 11 19,056.9 17.8% 17 28,454.2 26.6%

26 Rock Cliff 1026 1396 703,588.4 20% 140,717.7 18 59,862.

6 8.5% 44 114,028.8 16.2% 78 167,295.1 23.8%

27 Rock CliffE 1027 4314 1,768,132.6 20% 353,626.5 141 208,317

.7 11.8% 371 427,208.2 24.2% 562 575,892.1 32.6%

28 Rock CliffP 1028 11822 5,833,450.0 20% 1,166,690.0 231 497,884

.3 8.5% 582 1,035,866.8 17.8% 1,097 1,678,133.3 28.8%

29 Rock CliffVE 1029 218 125,300.5 20% 25,060.1 2 24,268.

5 19.4% 2 24,268.5 19.4% 4 27,505.7 22.0%

30 Rock CliffVP 1030 68 31,122.4 20% 6,224.5 1 2,207.6 7.1% 3 4,875.0 15.7% 5 6,344.0 20.4%

31 Rock Platform 1031 124 33,573.1 20% 6,714.6 4 5,216.8 15.5% 4 5,216.8 15.5% 8 7,930.7 23.6%

32 Rock PlatformE 1032 1774 661,917.3 20% 132,383.5 50 65,975.

0 10.0% 136 159,003.5 24.0% 201 214,936.5 32.5%

33 Rock PlatformP 1033 447 128,569.6 20% 25,713.9 12 13,167.

2 10.2% 23 22,847.2 17.8% 39 33,848.5 26.3%

34 Rock PlatformVE 1034 345 162,097.7 20% 32,419.5 2 18,431.

6 11.4% 2 18,431.6 11.4% 5 30,694.9 18.9%

35 Rock with Gravel Beach 1035 710 287,587.8 20% 57,517.6 12

19,232.9 6.7% 30 42,396.7 14.7% 68 83,143.3 28.9%

36 Rock with Gravel BeachE 1036 3374 1,394,014.4 20% 278,802.9 141

216,857.4 15.6% 340 436,901.9 31.3% 517 576,347.9 41.3%



Page 149

Case Target Target

_id Shore units


Goal amount

High 10% count

High 10%

length10% captured

Medium 20% count

Medium 20% length

20% captured

Low 30% count

Low 30% length

30% captured

37 Rock with Gravel BeachP 1037 9679 4,406,751.8 20% 881,350.4 246

385,559.8 8.7% 544 742,583.2 16.9% 1,064 1,282,102.0 29.1%

38 Rock with Gravel BeachVE 1038 123 40,373.1 20% 8,074.6 1 4,211.2 10.4% 2 8,161.7 20.2% 3 11,062.2 27.4%

39 Rock with Gravel BeachVP 1039 74 35,028.4 20% 7,005.7 3 5,272.1 15.1% 3 5,272.1 15.1% 5 7,500.6 21.4%

40 Rock with Sand & Gravel Beach 1040 173 57,913.7 20% 11,582.7 2 4,633.9 8.0% 7 12,650.8 21.8% 10 15,839.8 27.4%

41 Rock with Sand & Gravel BeachE 1041 2333 876,653.2 20% 175,330.6 92

143,711.1 16.4% 218 277,960.5 31.7% 327 366,484.6 41.8%

42 Rock with Sand & Gravel BeachP 1042 7600 2,843,714.0 20% 568,742.8 144

245,869.2 8.6% 348 487,979.6 17.2% 630 765,906.4 26.9%

43

Rock with Sand & Gravel BeachVE 1043 12 11,238.0 20% 2,247.6 1 2,330.9 20.7% 2 7,880.5 70.1% 2 7,880.5 70.1%

44

Rock with Sand & Gravel BeachVP 1044 39 15,471.3 20% 3,094.3 1 1,458.7 9.4% 3 3,035.9 19.6% 6 4,842.0 31.3%

45 Rock with Sand Beach 1045 19 4,346.8 20% 869.4 2 1,011.4 23.3% 4 1,358.5 31.3% 7 2,848.3 65.5%

46 Rock with Sand BeachE 1046 363 124,830.9 20% 24,966.2 17

22,770.0 18.2% 35 39,601.5 31.7% 45 48,338.4 38.7%

47 Rock with Sand BeachP 1047 544 182,404.5 20% 36,480.9 12

19,289.4 10.6% 21 28,422.3 15.6% 36 43,128.0 23.6%

48 Sand & Gravel Beach 1048 173 59,406.9 20% 11,881.4 3 8,605.2 14.5% 5 9,510.8 16.0% 11 14,823.6 25.0%

49 Sand & Gravel BeachE 1049 198 53,307.5 20% 10,661.5 5 5,060.4 9.5% 15 13,005.6 24.4% 23 17,342.6 32.5%

50 Sand & Gravel BeachP 1050 3600 1,339,292.9 20% 267,858.6 81

141,001.9 10.5% 179 255,142.5 19.1% 285 355,460.4 26.5%

51 Sand & Gravel BeachVP 1051 29 8,876.9 20% 1,775.4 3 2,046.9 23.1% 4 2,417.8 27.2% 4 2,417.8 27.2%

52 Sand & Gravel 1052 96 38,497.7 20% 7,699.5 3 3,393.4 8.8% 9 7,878.0 20.5% 14 11,377.9 29.6%



Page 150

Case Target Target

_id Shore units


Goal amount

High 10% count

High 10%

length10% captured

Medium 20% count

Medium 20% length

20% captured

Low 30% count

Low 30% length

30% captured

Flat

53 Sand & Gravel FlatE 1053 176 99,267.0 20% 19,853.4 13

17,622.1 17.8% 23 26,674.3 26.9% 34 35,132.2 35.4%

54 Sand & Gravel FlatP 1054 2689 1,298,001.0 20% 259,600.2 65

118,478.0 9.1% 173 251,221.4 19.4% 277 362,535.6 27.9%

55 Sand & Gravel FlatVP 1055 29 12,829.3 20% 2,565.9 1 1,502.8 11.7% 5 2,694.0 21.0% 6 3,293.6 25.7%

56 Sand Beach 1056 19 10,393.3 20% 2,078.7 1 3,729.3 35.9% 3 4,250.5 40.9% 6 6,759.0 65.0%

57 Sand BeachE 1057 242 187,660.8 20% 37,532.2 5 10,508.

3 5.6% 17 35,759.7 19.1% 29 53,661.0 28.6%

58 Sand BeachP 1058 298 124,058.7 20% 24,811.7 6 11,313.

4 9.1% 18 24,569.5 19.8% 31 38,185.7 30.8%

59 Sand Flat 1059 42 23,653.5 20% 4,730.7 2 2,764.4 11.7% 5 5,379.0 22.7% 9 11,549.1 48.8%

60 Sand FlatE 1060 129 78,238.5 20% 15,647.7 6 10,097.

1 12.9% 16 17,054.1 21.8% 24 22,162.8 28.3%

61 Sand FlatP 1061 2179 1,306,395.7 20% 261,279.1 37 126,694

.3 9.7% 84 234,122.0 17.9% 148 354,539.1 27.1%

62 Sand FlatVP 1062 40 22,440.9 20% 4,488.2 1 3,052.7 13.6% 5 10,008.9 44.6% 6 10,589.9 47.2%

63 Undefined (not a target) 9999 1026 662,992.3 0% 0.0 0 0.0 0.0% 2 18,830.3 2.8% 2 18,830.3 2.8%

Totals 59 TARGETS 61,095 27,340,266.0 5,468,053.2 1,514 2,748,7

18.6 3,634 5,540,862.2 6,217 8,297,748.1

Total shoreline contains 62,441 units with 28,145,005.1 m.

E = EXPOSED

P = PROTECTED

VE = VERY EXPOSED

VP = VERY PROTECTED




Page 151

Table 6.4 Fine Filter targets captured by nearshore Options A, B, and C

Target Target_id

Total Available Goal

Goal Amount (m)

High (10%)

10% captured

Medium (20%)

20% captured

Low (30%)

30% captured

Habitat3 1063 3,674,186.6 30% 1,102,256.0 668,967 18.2% 1,292,947 35.2% 1,655,988 45.1%

Eelgrass 1064 2,780,418.1 30% 834,125.4 545,632 19.6% 906,371 32.6% 1,171,633 42.1%

Surfgrass 1065 2,092,980.4 30% 627,894.1 392,045 18.7% 742,568 35.5% 966,681 46.2%

Saltmarsh 1066 2,669,922.6 30% 800,976.8 468,889 17.6% 861,477 32.3% 1,145,705 42.9%

Kelp 1067 5,134,530.6 30% 1,540,359.2 975,450 19.0% 1,663,440 32.4% 2,219,051 43.2%

Herring Spawning 1068 418,717.1 30% 125,615.1 96,838 23.1% 155,035 37.0% 204,617 48.9%

Thick-billed Murre seabird colony 1075 8,654.3 30% 2,596.3 5,583 64.5% 7,153 82.7% 7,153 82.7%

Cassin’s Auklet seabird colony 1074 322,925.4 30% 96,877.6 56,341 17.4% 116,426 36.1% 144,476 44.7%

Common Murre seabird colony 1073 18,896.5 30% 5,669.0 7,483 39.6% 9,053 47.9% 9,053 47.9%

Ancient Murrelet seabird colony 1072 251,716.0 30% 75,514.8 51,487 20.5% 82,559 32.8% 106,600 42.3%

Brandt’s Cormorant seabird colony 1071 5,296.6 30% 1,589.0 1,900 35.9% 1,900 35.9% 1,900 35.9%

Horned Puffin seabird colony 1070 15,476.0 30% 4,642.8 3,265 21.1% 6,520 42.1% 7,794 50.4%

Tufted Puffin seabird colony 1069 99,504.6 30% 29,851.4 24,833 25.0% 41,988 42.2% 47,069 47.3%



Page 152

6.3 Offshore Marine Spatial Analysis Results

Rather than just examining one set of model parameters, we have chosen instead to look at a range of different reserve sizes and a range of reserve fragmentation. From these, we then examined the results for emergent trends. Thus, rather than debating what is the “right” percentage to set aside, or whether larger reserves are better than several smaller ones, we have hopefully avoided these arguments for the time being by focusing on those areas that emerge under a variety of conditions. Those areas that were selected repeatedly we interpret as having a high “utility;” that is, usefulness, to marine reserve network design. While not necessarily meeting all goals, these areas of high overlap give clear direction as to where initial conservation efforts should be focused.

6.3.1 24 Scenarios; 2,400 Solutions

We examined 6 reserve network sizes: 5%, 10%, 20%, 30%, 40%, and 50%. In addition, we examined four MARXAN clumping parameters: very scattered, scattered, moderate, and moderately clumped (BLM = 0.0625, 0.250, 1.00, 4.00). For each of these 24 combinations of variables (6 reserve sizes x 4 clumpings), we ran MARXAN 100 times. Thus, we examined a total of 2,400 MARXAN solutions. For each of those 2,400 solutions, the algorithm performed 15 million iterations.

6.3.2 Utility

By looking at how many times a particular planning unit is included in a solution, we can get an indication of its utility in overall reserve network design. That is, those hexagons that are repeatedly chosen likely represent areas that are more useful for effective and efficient MPA network design. While it has been suggested that these hexagons may be “irreplaceable,” we have avoided using this terminology for two reasons:

1. This may cause some confusion with the irreplaceability heuristic which is part of the MARXAN software package, and is based on a completely different set of assumptions (Pressey et al. 1994, cited in MARXAN v1.8).

2. We are not actually saying that these areas are irreplaceable. While this may be true for some sites that harbour rare species (such as the hexactinellid sponge reefs), it is not necessarily so for all sites. Rather, these areas of high utility represent places that appear to be the most useful in the development of optimal reserve network solutions that best approach our targets, using a minimum of area. Less optimal solutions could possibly be found using larger areas of lower utility.

We have indicated the sum total of these 2,400 solutions as shades of blue (seldom chosen) to yellow (chosen frequently) in Map 45. The examination of various clumping values indicates that regardless of whether reserves are many and small, or few and large, certain areas recur over the course of many runs. For example, within the Central Coast, the following larger areas of high conservation utility emerge:

• Hexactinellid Sponge Reefs

• Goose Islands, Bardswell Islands, and vicinity



Page 153

• Rivers Inlet

• Scott Islands

• Entrance to Queen Charlotte Strait

• Broughton Archipelago

• Head of Knight Inlet

• Cordero Channel

While these areas alone would not constitute a fully representative Central Coast conservation portfolio, it is very likely that were they not included, such a portfolio would be difficult or impossible to achieve. Thus, regardless of what exact percentages were chosen by whatever planning processes, and the exact shape of the boundaries, we would expect the bright yellow areas to be key components of most conservation planning.

Larger areas of high conservation utility within the North Coast include:

• Hexactinellid Sponge Reefs

• West Aristazabal Island (and NW Price Island)

• Kitimat Arm

• Anger Island and vicinity

• SW and N Porcher Island, and Kitkatla Inlet

• S. Chatham Sound

• Mouth of Nass River

Larger areas of high conservation utility within the Haida Gwaii waters include:

• W. Dixon Entrance

• Naden Harbour

• Masset Inlet

• Skidegate Inlet (Kagan Bay)

• South Moresby Island

Larger areas of high conservation utility off N west coast Vancouver Island include:

• Scott Islands

• Mid-Quatsino Sound

• Brooks Peninsula (Cape Cook) westward to the base of the continental slope



Page 154

6.3.3 Flexible Solutions

Areas of high conservation utility alone would not constitute a fully representative conservation portfolio. The individual network solutions produced by Marxan can be diverse. Such diversity allows for greater flexibility when considering external factors, such as user interests, parks, local politics, and access and enforcement.

Once an initial selection of conservation areas has been chosen, probably based on the areas of high utility, but also taking into account the needs of the communities and stakeholders, the Marxan algorithm can be re-run, locking these areas into the network. Areas required to complete the portfolio (i.e., meeting the agreed-upon conservation goals) can then be explored.



Page 155

7.0 NEXT STEPS AND FUTURE WORK

Substantial work remains to be done with regard to regional analysis of biodiversity priorities and management in coastal BC. The authors strongly urge that improvements be made through the integration of this regional analysis into the work of the proposed coastal Ecosystem-Based Management Council that is being negotiated as part of the LRMP and ongoing government-to-government process.

7.1 Conservation Assessment of LRMP and LUP and Government-to-Government Decisions

As decisions are agreed to by planning tables and implemented through government-to-government decisions, it is important that the contribution of protected areas toward meeting the “low-risk” EBM goal is evaluated. Next steps should also include analysis of connectivity between protected areas, and an examination of their size, configuration and management. Such work is critical for exploring questions related to the adequacy of protected areas and management regimes for ensuring the long-term viability of the species and systems represented within. In particular, habitats, ecosystems, special elements, and connectivity zones that are not well represented by protected areas, need to be identified and targeted for high levels of protection through EBM.

7.2 Special Elements Refinement

Further work is required to bring together comprehensive occurrence information relative to special element targets. A series of small workshops with local experts, designed to identify such locations would greatly enhance the utility of any regional analysis in an ecosystem-based management framework.

7.3 Integrated Marine, Terrestrial, Freshwater

Efforts should be made to integrate nearshore, offshore and terrestrial/marine assessments into a single analysis for the purposes of optimizing conservation solutions between these environments. Such integration would also allow for important process linkages between environments to be characterized. In particular, accounting for integrated conservation along the land/sea interface is necessary so that associated uplands are protected along with high conservation priority shorelines. Likewise, it is important that estuaries and high priority freshwater systems are conserved in tandem.

7.4 Alignment of Terrestrial Classification Systems

Several approaches for assessing terrestrial ecosystem representation have been proposed and employed throughout the work of the CIT. These combinations have included the two approaches described in this report, as well as the development of small scale Predictive Ecosystem Mapping (ssPEM), and analysis units by BEC Variant combination used for work on



Page 156

EBM. The use of a single, nested hierarchical classification, based on existing provincial classification schemes, and common to multiple scales of analysis, is of particular importance for the future application of EBM and for the integration of regional scale objectives, targets and thresholds to planning and monitoring work at finer scales.

7.5 Enhancements to Salmon Modelling

While the ESA has applied an innovative approach to assessing the conservation status of salmon populations using existing escapement data, there are agreed upon shortcomings in this data set, and the clear need for supplemental information. Further research is needed to better assess stock area boundaries and conservation units. More rigorous, standardized and geographically extensive surveys are also required to improve upon the limited nature of current escapement data. Finally, research into the relationship between salmon population status and watershed condition would enhance our understanding of the role on-the-ground management regimes play with regard to salmon viability.

7.6 Species Penalties/Variable Goals

Given limitations in time, the ESA applied a simple rule for each SITES run that assigned equal representation goals to all ecosystems (expressed as a percentage of available area) and focal species (expressed as percentage of total available habitat score). Further, the species penalty objective was not used in SITES to differentiate priorities for meeting representation goals in the analysis. Future work should look at tailoring goals more specifically to species and ecosystems based on known habitat requirements, population viability assessments (PVAs), species/system distribution, and conservation status.

7.7 Temporal Modelling

One of the chief concerns arising from reviewers of the ESA’s methodological approach is the lack of temporal dynamic modelling relevant to forest ecosystems. Time and resources did not allow for inclusion of such an analysis, but future refinements of the ESA should examine the application of models such as SELES at a regional scale for ecosystem targets.

7.8 Cross-Scale Modelling

A key challenge for any ecosystem-based management framework emerging from LRMP negotiations will be the integration of analyses across multiple scales. Clearly the regional scale assessment presented here needs to be informed by analysis at sub-regional, landscape, watershed, and stand-level analysis. Likewise, these analyses at finer scales must be linked with information at the regional scale, and in particular, should be rolled-up for the purposes of creating regional-scale measures that feed into a study-area wide monitoring regime.



Page 157

8.0 CONCLUSIONS

The CIT ESA constitutes the first regional biodiversity analysis of its kind for BC’s Central Coast, North Coast and Haida Gwaii. Our purpose was to assess conservation values for terrestrial, freshwater, and marine ecosystems and species to provide information and analysis to support decision-making regarding regional land use.

Substantial challenges were faced in the production of the ESA, not the least of which involved data and technical limitations posed by undertaking a planning initiative for such a large study area. In particular, the availability of data sets that were regionally comprehensive limited the ability of the team to address special elements within the study. Likewise, models for focal species and ecosystems must be recognized as regional in scale, and the information is not appropriate, or intended for, decision-making at stand or operational scales. Further, limitations in time, human resources and financial resources required the ESA to limit its analysis to a static assessment of conservation values, as they currently exist on the landscape.

Despite these limitations, the ESA does represent a critical first step towards building an information base that can be used for making decisions about biodiversity conservation requirements on the coast of BC. The information and models developed to describe focal species habitats, as well as terrestrial, freshwater and marine ecosystems, represent the only existing synthesis of such information for the region. This synthesis forms the basis for prioritizing the region for conservation attention, and provides decision-makers with an assessment tool that evaluates both current and proposed protected areas scenarios.

While much debate continues regarding thresholds for sufficiency for species and ecosystems (i.e., how much is enough?), the ESA clearly demonstrates that most ecosystem types native to the study area are not currently represented in existing protected areas. As this report is being written, decisions are being made by planning tables that will substantially improve this situation. As these proposed protected areas are being negotiated, the ESA should be used to identify areas of highest conservation priority. Ideally, with such information in hand, protected areas will be situated in locations overlapping with Tier One and Two conservation areas identified by the ESA. On a more practical level, the ESA team recognizes that negotiations around protected areas involve a sequential, step-wise approach in which trade-offs are made between competing ecological, cultural, and economic interests. Therefore, the ESA has also provided maps and data sets describing a Relative Biodiversity Index, which specifically addresses the needs of stakeholders in such a decision-making environment.

Once protected areas have been agreed upon by planning tables, efforts must be focused on using the ESA to assess how well species and ecosystems are represented, and information emerging from gap analysis for all three LRMP areas needs to be highlighted for future work around ecosystem-based management. Examples of habitats and systems not well represented need to be identified, and well protected through additional landscape, watershed and stand-level reserves and prescriptive management guidelines. It is also important to note that even those species and systems well represented within watersheds designated for protection may be inadequately conserved if protected areas are not well designed and managed. For example, continuation of commercial trophy hunting and mineral exploration/extraction might seriously impair the ability of these areas to maintain species and system viability. Next steps should include analysis of



Page 158

connectivity between proposed protected areas, and an examination of their size, configuration and management. Such work is critical for exploring questions related to the adequacy of proposed protected areas for ensuring the long-term viability of the species and systems represented within.

While the lack of a comprehensive viability analysis for ESA species and systems prevents the identification of specific thresholds regarding sufficiency, previous studies with similar species and systems targeted by the ESA (but in dissimilar geographies), have identified representation requirements that fit within benchmarks established by the CIT’s EBM Planning Handbook. According to the Handbook, if the study area is to be managed for low ecological risk, 70% of the range of natural variation of habitats and ecosystems needs to be represented and conserved regionally. EBM provides for a number of ways for such goals to be achieved at various geographic scales, but from a regional perspective, one of the most certain ways to carry out such conservation is to institute “no risk” management regimes to capture a significant proportion of that 70% representation goal. The authors of this report acknowledge the uncertainties surrounding recommendations regarding the exact percentage of the study area that needs to be set aside in parks and protected areas. Nonetheless, we cannot escape the conclusion that more protection is better than less, and that more certain protection (e.g., park designation) is better than less certain protection (e.g., results-based regulation).

Substantial work remains to be done with regard to regional analysis of biodiversity priorities and management in coastal BC. The authors strongly urge that improvements be made through the integration of the ESA regional analysis into the proposed ongoing analysis to be undertaken by the coastal Ecosystem-Based Management Council that is being negotiated as part of the LRMP process. Future efforts must be directed at analyzing the outcomes of LRMP and LUP negotiations and at bringing new research and data into the analysis, including robust viability assessments and the addition of temporal modelling of dynamic ecosystem processes at the regional scale. Much work also remains to be done in terms of integrating marine and terrestrial analysis. Further, the regional scale analysis described in this report must be fully integrated into an EBM framework that accounts for conservation at finer scales—and vise versa—stand and operational management plans must be rolled-up and assessed across the region to provide a full assessment of the cumulative effect of resource management decisions.



Page 159

9.0 LITERATURE CITED

Adams, M.J. and R.B. Bury. 2000. Association of stream amphibians with climate gradients and the characteristics of headwater streams. Pp. 25-31 in R.B. Bury and M.J. Adams (editors). Inventory and monitoring and amphibians in North Cascades and Olympic National Parks, 1995-1998. Final Report to the Forest and Rangeland Ecosystem Science Center, U.S. Department of Interior and U.S. Geological Survey. Alidina, H. In Review. Gulf of Maine/Scotian Shelf MPA Planning Project. Selecting Priority Areas for Conservation, Marxan Analyses, Version 1: July 30, 2003. Technical Draft for Comment. WWF-Canada and CLF. Andelman, S.A., I. Ball, and D. Stomms. 1999. Sites 1.0: An Analytical Toolbox for Designing Ecoregional Conservation Portfolios. The Nature Conservancy, Arlington, Va. Archibald, W.R., R. Ellis, and A.N. Hamilton. 1987. Responses of grizzly bears to logging truck traffic in the Kimsquit River valley, British Columbia. Ursus 7:251-257 Ardron, J.A. 2002. A Recipe for Determining Benthic Complexity: An Indicator of Species Richness. Chapter 23, Marine Geography: GIS for the Oceans and Seas. (J. Breman, editor). ESRI Press, Redlands, Calif. Armstrup, S.C. and J. Beecham. 1976. Activity patterns of radio collared black bears in Idaho. Journal of Wildlife Management 40:340-348. Association for Biodiversity Information (ABI). 2001, Preparation for Biotics and Implementation of the Revised EO Methodology, Biotics Team Information Division. AXYS. 2001. British Columbia Marine Ecological Classification Update. Final Report. Submitted to Land Use Coordination Office, by AXYS Environmental Consulting Ltd., in association with John Roff, Ellen Hines. Bailey, R.G. 1998. Ecoregions: the ecosystem geography of the oceans and continents. Springer-Verlag, New York, N.Y. 176 p. Ballantine, W.J. 1997. Design principles for systems of "No-Take" Marine Reserves. Keynote Paper, from The Design & Monitoring of Marine Reserves. Fisheries Centre Research Reports, Vol. 5, No. 1. Banfield, A.W.F. 1974. The mammals of Canada. University of Toronto Press, Toronto, Ont. 438 p. Banner, A., R.J. Hebda, E.T. Oswald, J.Pojar, and R. Trowbridge. 1988. Wetlands of Pacific Canada. Pp. 306-346 in Wetlands of Canada, National Wetlands Working Group. Ottawa, Ont. Polyscience. Banner, A., W. MacKenzie, S. Haeussler, S. Thompson, J. Pojar, and R. Trowbridge. 1993. A field guide to site identification and interpretation for the Prince Rupert Forest Region. B.C. Ministry of Forests, Research Branch, Victoria, B.C.



Page 160

Banner, A., J. Pojar, and R. Trowbridge. 1986. Representative Wetland Types of the Northern Part of the Pacific Oceanic Wetland Region. B.C. Min. For. Res. Rep., Smithers, B.C. 45 p. Baron, J.S., N.L. Poff, P.L. Angermeier, C.N. Dahm, P.H. Gleick, N.G. Hairston, Jr., R.B. Jackson, C.A. Johnston, B.G. Richter, and A.D. Steinman. 2002. Meeting ecological and societal needs for freshwater. Ecological Applications 12:1247-1260. Beck, M. W., M. Odaya, J. J. Bachant, J. Bergan, B. Keller, R. Martin, R. Mathews, C. Porter, and G. Ramseur. 2000. Identification of Priority Sites for Conservation in the Northern Gulf of Mexico: An Ecoregional Plan. The Nature Conservancy, Arlington, Va. Bedward, M., R.L. Pressey, and D.A. Keith. 1992. A new approach for selecting fully representative reserve networks: addressing efficiency, reserve design and land suitability with an iterative analysis. Biological Conservation 62: 115-125. Beissinger, S.R. and D.R. McCullough (editors). 2002. Population viability analysis. University of Chicago Press, Chicago, Ill. Ben-David, M., T.A. Hanley, and D.M. Schell. 1998. Fertilization of terrestrial vegetation by spawning Pacific salmon: the role of flooding and predator activity. Oikos 83:47-55. Benn, B. 1998. Grizzly bear mortality in the Central Rockies Ecosystem. MSc. thesis. Faculty of Environmental Design, University of Calgary. 151 p. plus appendices. Berry, H.D., J.R. Harper, T.F. Mumford, Jr., B.E. Bookheim, A.T. Sewell, and L.J. Tamayo. 2001. The Washington State shorezone inventory user’s manual. Nearshore Habitat Program, Washington State Department of Natural Resources, Olympia, Wash. Bilby, R.E., B.R. Fransen, P.A. Bisson, and J.K. Walter. 1998. Response of juvenile coho salmon (Oncorhynchus kisutch) and steelhead (Oncorhynchus mykiss) to the addition of salmon carcasses to two streams in southwestern Washington, U.S.A. Canadian Journal of Fisheries and Aquatic Sciences 55:1909-1918. Booth, J., Coastal and Ocean Resources Inc., and Clover Point Cartographics. 1998. Study to Identify Preliminary Representative Marine Areas in the Queen Charlotte Sound Marine Region. Report prepared for National Parks, Parks Canada. Boulanger, J. and S. Himmer. 2001. Kingcome (1997) DNA mark-recapture grizzly bear inventory project. Final Report. B.C. Ministry of Environment, Lands and Parks, Nanaimo. 83 p. Boyce, M.S. 1992. Population viability analysis. Annual Review of Ecology and Systematics 23:481-506. Boyce, M.S. and L.L. McDonald. 1999. Relating populations to habitats using resource selection functions. Trends in Ecology and Evolution 14:268-272.



Page 161

Brashares, J. S., P. Arcese, and M. K. Sam. 2001. Human demography and reserve size predict wildlife extinction in West Africa. Proceedings of the Royal Society of London Series B-Biological Sciences 268:2473-2478. Briers, R.A. 2002. Incorporating connectivity into reserve selection procedures. Biological Conservation 103:77-83. British Columbia Conservation Data Centre (BC CDC) Web page. “Endangered Species and Ecosystems in British Columbia.” URL: http://srmwww.gov.bc.ca/atrisk/toolintro.html British Columbia Conservation Data Centre (BC CDC). 2003. Conservation Data Centre Element Occurrence Records digital datafiles. B.C. Ministry of Sustainable Resource Management. British Columbia Conservation Data Centre (BC CDC). 2003. Conservation Data Centre Sensitive Element Occurrence Records digital datafiles. B.C. Ministry of Sustainable Resource Management. British Columbia Land Use Coordination Office (LUCO). 1997. British Columbia marine ecological classification: Marine ecosections and ecounits. Version 1.0. Land Use Coordination Office, Otter Bay Coastal Resources, and Coastal and Oceans Resources Inc. for the Coastal Task Force, Resources Inventory Committee. British Columbia Ministry of Environment, Lands and Parks, and BC Ministry of Forests. 2001. Inventory methods for tailed frog and Pacific giant salamander. Standards for Components of British Columbia’s Biodiversity No. 39. Resource Inventory Branch, MOELP, and Resource Branch, MOF, Victoria, B.C. British Columbia Ministry of Environment, Lands and Parks (MELP). 1995. A future for the grizzly: British Columbia Grizzly Bear Conservation Strategy. Victoria, B.C. 16 p. British Columbia Ministry of Forests and B.C. Ministry of Environment. 1999. Managing Identified Wildlife Guidebook, Volume I. Victoria, B.C. Brosofke, K.D., J. Chen, R.J. Naiman, and J.F. Franklin. 1997. Harvesting effects on microclimatic gradients from small streams to uplands in western Washington. Ecological Applications 7:1188-1200. Brown, H.A. 1975. Temperature and development of the tailed frog, Ascaphus truei. Comparative Biochemistry and Physiology, Great Britain 50:397-405. Brown, L.E., D.M. Hannah, and A.M. Milner. 2003. Alpine stream habitat classification: An alternative approach incorporating the role of dynamic water source contributions. Arctic, Antarctic, and Alpine Research 35(3):313-322. Burger, A. E. and T.A. Chatwin (editors). 2002. Multi-scale studies of populations, distribution and habitat associations of Marbled Murrelets in Clayoquot Sound, British Columbia. Ministry of Water, Land and Air Protection, Victoria, B.C. 170 p.



Page 162

Bury, R.B., P. Loafman, D. Rofkar, and K.I. Mike. 2001. Clutch sizes and nests of tailed frogs from the Olympic Peninsula, Washington. Northwest Science 75(4):419-422. Calder, J. and R. Taylor. 1968. Flora of the Queen Charlotte Islands. Monograph No. 4, part 1. Canada Dept. of Agriculture, Plant Research Institute. Ottawa, Ont. Campbell, R.W., N.K. Dawe, I. McTaggart-Cowan, J.M. Cooper, G.W. Kaiser, and M.C.E. McNall. 1990. The birds of British Columbia, Volumes 1 - IV. UBC Press, Vancouver, B.C. Canadian Wildlife Service. 2001. British Columbia Seabird Colony Inventory: Digital dataset. Canadian Wildlife Service, Pacific and Yukon Region, British Columbia. Canadian Wildlife Service. 2002. Critical Waterfowl Habitat in British Columbia. Report and digital dataset. Canadian Wildlife Service, Pacific and Yukon Region, British Columbia. Cannings, R. and S. Cannings. 1996. British Columbia: A natural history. Greystone Books, Douglas & McIntyre, Vancouver/Toronto. Cannings, S.G. and J. Ptolemy. 1998. Rare Freshwater Fish of British Columbia. B.C. Ministry of Environment, Lands and Parks Wildlife Branch and B.C. Ministry of Forests Research Branch, Victoria, B.C. 214 p. Cannings, S.G., L.R. Ramsay, D.F. Fraser, and M.A. Fraker. 1999. Rare Amphibians, Reptiles, and mammals of British Columbia. Ministry of Environment, Lands and Parks Wildlife Branch and Resources Inventory Branch, Victoria, B.C. 190 p. Carr, M.H. and D.C. Reed. 1993. Conceptual issues relevant to marine harvest refuges: examples from temperate marine fishes. Canadian Journal of Fisheries and Aquatic Sciences 50: 2019-2028 Carroll, C., R.F. Noss, and P.C. Paquet. 2001. Carnivores as focal species for conservation planning in the Rocky Mountain region. Ecological Applications 11:961-980. Carroll, C., R.F. Noss, P.C. Paquet, and N.H. Schumaker. 2003. Use of population viability analysis and reserve selection algorithms in regional conservation plans. Ecological Applications 13(6): 1773-1789. Cederholm, C.J., D.H. Johnson, R.E. Bilby, L.G. Dominguez, A.M. Garrett, W.H. Graeber, E.L. Greda, M.D. Kunze, B.G. Marcot, J.F. Palmisano, R.W. Plotnikoff, W.G. Pearcy, C.A. Simenstad, and P.C. Trotter. 2001. Pacific salmon and wildlife: ecological contexts, relationships, and implications for management. Pp. 628-685 in D.H. Johnson and T.A. O’Neil (editors). Wildlife-habitat relationships in Oregon and Washington. Oregon State University Press, Corvallis, Oreg. Cederholm, C.J., D.B. Houston, D.L. Cole, and W.J. Scarlett. 1989. Fate of coho salmon (Oncorhynchus kisutch) carcasses in spawning streams. Canadian Journal of Fisheries and Aquatic Sciences 46:1347-1355. Chadwick, D.H. 1977. The influence of mountain goat social relationships on population size and distribution. Pp. 74-91 in W.G. M. Samuel (editor). Proceedings, First international mountain goat



Page 163

symposium. Province of British Columbia Ministry of Recreation and Conservation, Fish and Wildlife Branch, Kalispell, Mont. Chen J., J.F. Franklin, and T.A. Spies. 1995. Growing season microclimatic gradients from clearcut edges into old-growth Douglas-fir forests. Ecological Applications 5:74-86. Chen, J., J.F. Franklin, and T.A. Spies. 1993. Contrasting microclimates among clearcut, edge, and interior of old-growth Douglas-fir forest. Agri. For. Metero. 63:219-237. Church, R., R. Gerrard, A. Hollander, and D. Stoms. 2000. Understanding the tradeoffs between site quality and species presence in reserve site selection. Forest Science 46(2):157-167. Claussen, D.L. 1973. The water relations of the tailed frog, Ascaphus truei and the Pacific treefrog, Hyla regilla. Comparative Biochemistry and Physiology. Peragmon Press, U.K. Comer, P. 2001. The Nature Conservancy, Conservation Science Division Memorandum, Re: Observations and recommendations for setting conservation goals in ecoregional plans, January 8, 2001. The Nature Conservancy, Conservation Science Division, Boulder, Colo. 11 p. Committee On the Status of Endangered Wildlife in Canada Web site. “Species At Risk Status Reports.” URL: http://www.sararegistry.gc.ca/status/status_e.cfm (Accessed May 2003). Connolly, S.R. and J. Roughgarden. 1997. A latitudinal gradient in northeast Pacific intertidal community structure: Evidence for an oceanographically based synthesis of marine community theory. American Naturalist 155:4. Conway, K.W. 1999. Hexactinellid Sponge Reefs on the British Columbia continental shelf: geological and biological structure with a perspective on their role in the shelf ecosystem. Canadian Stock Assessment Secretariat Research Document. 99/192. Conway, K.W., M. Krautter, J.V. Barrie, and M. Neuweiler. 2001. Hexactinellid sponge reefs on the Canadian shelf: a unique “living fossil.” Geoscience Canada 28(2):71-78. Cooper, J.M. and V. Stevens. 2000. A Review of the Ecology, Management and Conservation of the Northern Goshawk in British Columbia. B.C. Ministry of Environment, Lands and Parks, Wildlife Bulletin No. B-101. Corn, P.S. and R.B. Bury. 1989. Logging in western Oregon: responses of headwater habitats and stream amphibians. Forest Ecology and Management 29:39-57. Cowling, R.M., R.L. Pressey, M. Rouget, and A.T. Lombard. 2003. A conservation plan for a global biodiversity hotspot – the Cape Floristic Region, South Africa. Biological Conservation 112:191-216. Cox, J., R. Katz, M. MacLaughlin and T. Gilbert. 1994. Closing the gaps in Florida’s Wildlife Habitat Conservation System. Tallahassee: Florida Game and Freshwater Fish Commission Report.



Page 164

Crawford, W.R. and R.E. Thomson. 1991. Physical oceanography of the western Canadian continental shelf. Continental Shelf Research 11 (8-10):669-683. Csuti, B., S. Polasky, P.H. Williams, R.L. Pressey, J.D. Camm, M. Kershaw, A.R. Kiester, B. Downs, R. Hamilton, M. Muso, and K. Sahr. 1997. A comparison of reserve selection algorithms using data on terrestrial vertebrates in Oregon. Biological Conservation 80:83-97 Cumming, S.G., P.J. Burton, and B. Klinkenberg. 1996. Boreal mixedwood forests may have no “representative” areas: Some implications for reserve design. Ecography 19(2):162-180. Dale, N. 1997. An overview of key conservation, recreation and cultural heritage values in British Columbia’s marine environment. A report prepared for the Land Use Coordination Office, Ministry of Environment Lands and Parks, BC. Available online: URL: http://www.luco.gov.bc.ca/coastal/mpatoc.htm Darimont, C., K. Gerwing, P. Paquet, B. Reid, W. McCrory, and B. Cross. 2003. Preliminary modeling of deer winter range in Heiltsuk Territory on the Central Coast of British Columbia (January 2003). Prepared for The Heiltsuk Nation and Raincoast Conservation Society. Victoria, B.C. 10 p. plus maps. Darimont, C.T. and P.C. Paquet. 2000. The gray wolves (Canis lupus) of British Columbia’s coastal rainforests: Findings from Year 2000 pilot study and conservation assessment. Prepared for the Raincoast Conservation Society. Victoria, B.C. 62 p. Davis, H. 1996. Characteristics and selection of winter dens by black bears in coastal British Columbia. Thesis. Simon Fraser University, Burnaby, B.C. Day, J. and K. Lavoie. 1998. Scotian Shelf Case Study: Applying the WWF framework to determine Marine Natural Regions and Marine Representative Units (Draft). A report prepared for WWF Canada, Dept. of Fisheries and Oceans, and Parks Canada. Day, J.C. and J.C. Roff. 2000. Planning for representative Marine Protected Areas: A framework for Canada’s Oceans. Report prepared for World Wildlife Fund Canada, Toronto, Ont. Demarchi, M.W. 2003. Grizzly bear monitoring for the Greenville to Kincolith road project. 2002 Annual Report. LGL Limited and B.C. Ministry of Water, Land and Air Protection, Smithers, B.C. 54 p. Dethier, M.N. 1990. A marine and estuarine habitat classification system for Washington State. Washington Natural Heritage Program, Department of Natural Resources. Olympia, Wash. 56 p. DFO 2002. Dept. of Fisheries and Oceans. Herring Web site. Point data: http://www.pac.dfo-mpo.gc.ca/sci/herring/herspawn/cumulati.htm DFO 2003. Dept. of Fisheries and Oceans. Habitat and Enhancement Branch. Eulachon estuary point data: http://www-heb.pac.dfo-mpo.gc.ca/maps/themesdata_e.htm



Page 165

Dinerstein E., D.M. Olson, D.J. Graham, A.L. Webster, S.A. Primm, M.P. Bookbinder, and G. Ledec. 1995. A conservation assessment of the terrestrial ecoregions of Latin America and the Caribbean. World Bank, Washington, D.C. Doak, D. F. 1995. Source-sink models and the problem of habitat degradation: general models and applications to the Yellowstone grizzly bear. Conservation Biology 9:1370-1379. Douglas, G.W., D. Meidinger, and J.L. Penny. 2002. Rare Native Vascular Plants of British Columbia. BC Environment, Victoria, B.C. 358 p. Dupuis, L.A. 1999. Status report on the Tailed Frog, Ascaphus truei, in Canada. Committee on the Status of Endangered Wildlife in Canada, Ottawa, Ont. 34 p. Dupuis, L.A. and P. Friele. 1996. Riparian management and the tailed frog. Final report for Ministry of Forests, Prince Rupert Region. Smithers, B.C. 25 p. Dupuis, L.A. and P. Friele. 2002. Distribution of Ascaphus montanus in the Yahk River and neighbouring watersheds. Report to Tembec Industries, Cranbrook, B.C. and Columbia Basin Fish and Wildlife Compensation Program, Nelson, B.C. Dupuis, L.A. and P. Friele. 2003. Watershed-level protection and management measures for the maintenance of Ascaphus truei populations in the Skeena Region. Draft Report for Ministry of Water, Land and Air Protection, Skeena Region, Smithers, B.C. 54 p. Dupuis L.A. and J.D. Steventon. 1999. Riparian management and the Tailed Frog in northern coastal forests. Forest Ecology and Management 124:35-43. Eklov, P. 2000. Chemical cues from multiple predator-prey interactions induce changes in behavior and growth of anuran larvae. Oecologica 123: 192-199. Estes, J.A. 1990. Growth and equilibrium in sea otter populations. Journal of Animal Ecology 59:385-401. Estes, J.A., M.T. Tinker, T.M. Williams, and D.F. Doak. 1998. Killer whale predation on sea otters linking oceanic and nearshore ecosystems. Science 282: 473-475. Etnoyer, P. and L. Morgan. 2003. Occurrences of habitat-forming cold water corals in the Northeast Pacific Ocean (Draft May 2003). A report to NOAA’s Office of Habitat Protection. Marine Conservation Biology Institute, Redmond, Wash. Evenson, J.R. and J.B. Buchanan. 1997. Seasonal abundance of shorebirds at Puget Sound estuaries. Washington Birds 6:34-62. Fairbanks, D.H.K. and G.A. Benn. 2000. Identifying regional landscapes for conservation planning: a case study from KwaZulu-Natal, South Africa. Landscape and Urban Planning 50:1-19.



Page 166

Fairbanks, D.H.K., M.W.M. Thompson, D.E. Vink, T.S. Newby, H.M. van den Berg, and D.A. Everard. 2000. The South African land-cover characteristics database: a synopsis of the landscape. South African Journal of Science 96:69-82. Fargo, J. and A. Tyler. 1991. Sustainability of flatfish-dominated fish assemblages in Hecate Strait, British Columbia, Canada. Netherlands Journal of Sea Research 27(3/4):237-253. Fargo, J. and A. Tyler. 1992. Statistical testing of research trawl data with implications for survey design. Netherlands Journal of Sea Research 29(1-3):97-108. Findlay, C.S. and J. Houlahan. 1997. Anthropogenic correlates of biodiversity in southeastern Ontario wetlands. Conservation Biology 11:1000-1009. Finney, B.P., I. Gregory-Eaves, J. Sweetman, M.S.V. Douglas, and J.P. Smol. 2000. Impacts of climatic change and fishing on Pacific salmon abundance over the past 300 years. Science 290:795-799. Fishbase Web site. “Organisms Preying on Thaleichthys pacificus.” URL: http://www.fishbase.org/TrophicEco/PredatorList.cfm?ID=256&GenusName=Thaleichthys&SpeciesName=pacificus (Accessed Dec. 01). Ford, J.K.B., G.M. Ellis, L.G. Barrett-Lennard, A.B. Morton, R.S. Palm, and K.C. Balcolm. 1998. Dietary specialization in two sympatric populations of killer whales (Orcinus orca) in coastal British Columbia and adjacent waters. Canadian Journal of Zooogy 76:1456-1471. Forman, R. 1995. Land mosaics: The ecology of landscapes and regions. Cambridge University Press, Cambridge, U.K. Forman, R.T.T., D.S. Friedman, D. Fitzhenry, J.D. Martin, A.S. Chen, and L.E. Alexander. 1997. Ecological effects of roads: Toward three summary indices and an overview for North America. Pp. 40-54 in K. Canters, A. Piepers, and D. Hendriks-Heersma (editors). Habitat fragmentation, infrastructure and the role of ecological engineering. Proceedings of the International Conference. Ministry of Transport, Public Works and Water Management. Maastricht—The Hague. Delft, The Netherlands. September 17-21. Foster, B.R. 1982. Observability and habitat characteristics of the mountain goat (Oreamnos americanus Blainville, 1816) in west-central British Columbia. University of British Columbia, Vancouver, B.C. Fox, J.L. 1977. Summer mountain goat activity and habitat preferences in coastal Alaska as a basis for the assessment of survey techniques. Pp. 190-199 in W. Sammuel and. W.G. Macgregor (editors). Proceedings, First international mountain goat symposium. Province of British Columbia Ministry of Recreation and Conservation, Fish and Wildlife Branch, Kalispell, Mont. Fox, J.L. and C.A. Smith. 1988. Winter mountain goat diets in southeast Alaska [USA]. Journal of Wildlife Management 52:362-365. Fox, J.L., C.A. Smith, and J.W. Schoen. 1989. Relation between mountain goats and their habitat in southeastern Alaska. USDA Forest Service, General Tech. Rep PNW-GTR-246, Washington, D.C.



Page 167

Franklin, J.F., F. Hall, W. Laudenslayer, C. Maser, J. Nunan, J. Poppino, C.J. Ralph, and T. Spies. 1986. Interim definitions for old-growth Douglas-fir and mixed-conifer forests in the Pacific Northwest and California. Research Note PNW-447. USDA Forest Service, Washington, D.C. Franklin J.F.K., D.W. Cromack Jr, A. McKee, C. Maser, J. Sedell, F. Swanson, and G. Juday. 1981. Ecological characteristics of old-growth Douglas-fir forests. USDA Forest Service, General Technical Report PNW-118, Portland, Oreg. Fuhr, B. and D.A. Demarchi. 1990. A methodology for grizzly bear habitat assessment in British Columbia. Wildlife Bulletin No. B-67. B.C. Ministry of Environment, Victoria, B.C. 28 p. Gibeau, M.L., A.P. Clevenger, S. Herrero, and J. Wierzchowski. 2002. Grizzly bear response to human development and activities in the Bow River Watershed, Alberta, Canada. Biological Conservation 103(2):227-236 Gilbert, F.F. and R. Allwine. 1991. Terrestrial amphibian communities in the Oregon Cascade Range. Pp. 319-326 in L.F. Ruggiero, K.B. Aubry, A.B. Carey, and M.H. Huff (technical coordinators). Wildlife and vegetation of unmanaged Douglas-Fir forests. USDA Forest Service, Gen. Tech. Rep. PNW-GTR-285. Goward, T. 1999. The lichens of British Columbia, Illustrated Keys, Part 2 - Fruticose Species. Ministry of Forests Research Program. Victoria, B.C. 319 p. Gresh, T., J. Lichatowich, and P. Schoonmaker. 2000. An estimation of historic and current levels of salmon production in the Northeast Pacific ecosystem: Evidence of a nutrient deficit in the freshwater systems of the Pacific Northwest. Fisheries 25(1):15-21. Gross, J.E., M.C. Kneeland, D.F. Reed, and R.M. Reich. 2002. GIS-based habitat models for mountain goats. Journal of Mammalogy 83:218-228. Groves, C. 2003. Drafting a conservation blueprint: A practitioner’s guide to planning for biodiversity. Island Press, Washington, D.C. Groves, C., D.B. Jensen, L.L. Valutis, K.H. Redford, M.L. Shaffer, J.M. Scott, J.V. Baumgartner, J.V. Higgins, M.W. Beck, and M.G. Anderson. 2002. Planning for Biodiversity Conservation: Putting Conservation Science into Practice. Bioscience 52 (6): 499-512. Groves, C.R., L. Valutis, D. Vosick, B. Neely, K. Wheaton, J. Touval, and B. Runnels. 2000. Designing a geography of hope: A practitioner’s handbook for ecoregional conservation planning. 2nd ed., vols. 1 and 2. The Nature Conservancy, Arlington, Va. Haas, G. 1998. Indigenous fish species potentially at risk in British Columbia, with recommendations and prioritizations for conservation, forestry/resource use, inventory and research. B.C. Ministry of Fisheries, Fisheries Management Report No. 105. Halpern, B. 2003. The impact of marine reserves: Do reserves work and does size matter? Ecological Applications 13:1 supplement 117-137.



Page 168

Hamilton, A.N. 1987. Classification of coastal grizzly bear habitat for forestry interpretations and the role of food in habitat use by coastal grizzly bears. MSc. thesis, Faculty of Forestry, University of British Columbia. 83 p. Hamilton, A.N., W.R. Archibald, and E.C. Lofroth. 1986. Coastal grizzly research project. Progress report - year 4, Working plan - year 5. Wildlife Working Report WR-22. B.C. Ministries of Environment and Parks and Forests. 92 p. Hamilton, A.N. and M.A. Austin. 2002. Grizzly bear harvest management in British Columbia: Background report. B.C. Ministry of Water, Land and Air Protection, Victoria, B.C. 95 p. Harestad, A.S. 1985. Habitat use by black-tailed deer on northern Vancouver Island. Journal of Wildlife Management 49: 946-950. Harper, J.R. and H. Berry. 2000. User’s manual for the biophysical shore-zone mapping system of Washington. Washington Department of Natural Resources, Olympia, Wash. Harper, J.R., J. Christian, W.E. Cross, R. Frith, G. Searing, and D. Thomson. 1993. A classification of the marine regions of Canada, Final Report. Coastal & Oceans Resources, Inc., Sidney, B.C. Harris, R. B., and F. W. Allendorf. 1989. Genetically effective population size of large mammals: an assessment of estimators. Conservation Biology 3:181-191. Hart, D.D. and C. M. Finelli. 1999. Physical-biological coupling in streams: The pervasive effects of flow on benthic organisms. Annual Review of Ecology and Systematics 30: 363-395. Hart, J.L. 1973. Pacific fishes of Canada. Fisheries Research Board of Canada Bulletin 180. Hatler, D.F. 1967. Some aspects of the ecology of black bears (Ursus americanus) in Interior Alaska. M.S. thesis, University of Alaska College. 111 p. Hay, D.E. and P.B. McCarter. 2001. Herring spawn areas of British Columbia, A review, geographic analysis and classification. Fisheries and Oceans Canada, Pacific Biological Station, Nanaimo, British Columbia. Revised edition: 2001. URL: http://www.pac.dfo-mpo.gc.ca/sci/herring/herspawn/pages/acknowl1_e.htm (accessed Jan. 2002). Haynes, L. A. 1992. Mountain goat habitat of Wyoming’s Beartooth Plateau: Implications for management. Proc. of the Biennial Symposium of the Northern Wild Sheep and Goat Council 8:325-339. Hebert, D.M. and W.G. Turnbull. 1977. A description of southern interior and coastal mountain goat ecotypes in British Columbia. Pp. 126-146 in W. Sammuel and W.G. Macgregor (editors). Proceedings, First international mountain goat symposium. Province of British Columbia Ministry of Recreation and Conservation, Fish and Wildlife Branch, Kalispell, Mont. Helfield, J.M. and R.J. Naiman. 2001. Effects of salmon-derived nitrogen on riparian forest growth and implications for stream productivity. Ecology 82:2403-2409.



Page 169

Higgins, J.V., M.T. Bryer, M. Lammert, and T.W. Fitzhugh. 2003. A freshwater ecosystem classification approach for biodiversity conservation planning. BioScience (in prep.) Hilderbrand, G.V., S.D. Farley, C.T. Robbins, T.A. Hanley, K. Titus, and C. Servheen. 1996. Use of stable isotopes to determine diets of living and extinct bears. Canadian Journal of Zoology 74:2080-2088. Himmer, S. and J. Boulanger. 2003. Trends in grizzly bears utilizing salmon streams in the Owikeno Lake System: 1998-2002. Western Forest Products Ltd., Campbell River, and B.C. Ministry of Water, Land and Air Protection, Williams Lake, B.C. 28 p. Himmer, S. and M. Gallagher. 1996. Tweedsmuir grizzly bear study. Final report. B.C. Ministry of Environment, Lands and Parks, Victoria, B.C. 108 p. Hjeljord, O. 1973. Mountain goat forage and habitat preference in Alaska. Journal of Wildlife Management 37:353-362. Hocutt, C.H. and E.O. Wiley (editors). 1986. The zoogeography of North American freshwater fishes. John Wiley and Sons, New York, N.Y. 866 p. Hoctor, T.S., M.H. Carr, and P.D. Zwick. 2000. Identifying a linked reserve system using a regional landscape approach: the Florida ecological network. Conservation Biology 14(4):984-1000. Howes, D.E., J.R. Harper, and E.H. Owens 1994. Physical shore-zone mapping system for British Columbia. Technical Report by Coastal and Ocean Resources, Inc., for the Coastal Task Force of the Resource Inventory Committee (RIC), RIC Secretariat, Victoria, B.C. 71 p. Howes, D.E., P. Wainwright, J. Haggarty, J. Harper, E. Owens, D. Reimer, K. Summers, J. Cooper, L. Berg, and R. Baird. 1993. Coastal resources and oil spill response atlas for the southern Strait of Georgia. B.C. Ministry of Environment, Lands and Parks, Environmental Emergencies Coordination Office, Victoria, B.C. 317 p. Hunter, M.L., Jr. 1991. Coping with ignorance: the coarse-filter strategy for maintaining biodiversity. Pp. 266-281 in K. Kohm (editor), Balancing on the brink of extinction: the Endangered Species Act and lessons for the future. Island Press, Washington, D.C. Hunter, M.L., Jr., G.L. Jacobson, Jr., and T. Webb III. 1988. Paleoecology and the coarse filter approach to maintaining biological diversity. Conservation Biology 2:375-385. Iverson, G.C., G.D. Hayward, K. Titus, E. DeGayner, R.E. Lowell, D.C. Crocker-Bedford, P.F. Schempf, and J. Lindell. 1996. Conservation assessment for the northern goshawk in southeast Alaska. C.G. Shaw III, Technical Coordinator. Gen. Tech. Rep. PNW-GTR-387. USDA Forest Service, Pacific Northwest Research Station, Portland, Oreg. 101 p. Jamieson, G.S. and L. Chew. 2002. Hexactinellid Sponge Reefs: Areas of interest as Marine Protected Areas in the North and Central Coasts areas. Canadian Science Advisory Secretariat. Research document 2002/122.



Page 170

Joao, E.M. 1998. Causes and consequences of map generalisation. Taylor & Francis, London, U.K. Johnson, R.L. 1983. Mountain goats and mountain sheep of Washington. Biological Bulletin 18. Washington Dept. of Game, Olympia, Wash. Johnston, B. and L. Frid. 2002. Clearcut logging restricts the movements of terrestrial Pacific Giant Salamanders (Dicamptodon tenebrosus Good). Canadian Jouranl of Zoology 80:2170-2177. Jones, G.W. and F.L. Bunnell. 1984. Response of black-tailed deer to winters of different severity on northern Vancouver Island. Pp. 385-390 in W.R. Meehan, T.R. Merrell, Jr., and T.A. Hanley (editors). Fish and wildlife relationships in old-growth forests: proceedings of a symposium. Amer. Inst. Fish. Res. Biol. Reintjes Publ., Morehead City, N.C. Jones, J.A., F.J. Swanson, B.C. Wemple, and K.U. Snyder. 2000. Effects of roads on hydrology, geomorphology, and disturbance patches in stream networks. Conservation Biology 14 (1):76-85. Jones, J.A. and G.E. Grant. 1996. Long-term storm flow responses to clearcutting and roads in small and large basins, western Cascades, Oregon. Water Resources Research 32:959-974. Joslin, G. 1986. Montana mountain goat investigations, Rocky Mountain Front. Montana Dept. of Fish, Wildlife, and Parks, Helena, Mont. Jungwirth, M., S. Muhar, and S. Schmutz. 2002. Re-establishing and assessing ecological integrity in riverine landscapes. Freshwater Biology 47(4):867-887. Karr, J.R. 1991. Biological integrity: A long-neglected aspect of water resource management. Ecological Applications 1:66-84. Kintsch, J.A. and D.L. Urban. 2002. Focal species, community representation, and physical proxies as conservation strategies: as case study in the Amphibolite Mountains, North Carolina, U.S.A. Conservation Biology 16:936-947. Kirchhoff, M.D. and J.W. Schoen. 1987. Forest cover and snow: implications for deer habitat in southeast Alaska. Journal of Wildlife Management 51:28-33. Kirkpatrick, J.B. and M.J. Brown. 1994. A comparison of direct and environmental domain approaches to planning reservation of forest higher plant communities and species in Tasmania. Conservation Biology 8:217-224. Kline, T.C. Jr., J.J. Goering, O.A. Mathisen, P.H. Poe, and P.L. Parker. 1990. Recycling of elements transported upstream by runs of Pacific salmon: I. d15N and d13C evidence in Sashin Creek, southeastern Alaska. Canadian Journal of Fisheries and Aquatic Sciences 47:136-144. Krautter, M., K.W. Conway, J.V. Barrie, and M. Neuweiler. 2001. Discovery of a “living Dinosaur”: Globally unique modern hexactinellid sponge reefs off British Columbia, Canada. FACIES 44:265-282. Kuletz, K.J. and J.F. Piatt. 1999. Juvenile marbled murrelet nurseries and the productivity index. Wilson Bulletin 111(2):257-261.



Page 171

Lambeck, R.J. 1997. Focal species: A multi-species umbrella for nature conservation. Conservation Biology 2(4):849-856. Lande, R., S. Engen, and B. E. Saether. 1994. Optimal harvesting, economic discounting and extinction risk in fluctuating populations. Nature (London) 372:88-90. Landres, P.B., P. Morgan, and F.J. Swanson. 1999. Overview of the use of natural variability concepts in managing ecological systems. Ecological Applications 9:1179-1188. Leslie, H., M. Ruckelshaus, I. Ball, S. Andelman, and H. Possingham. 2002. Using siting algorithms in the design of marine reserve networks. Ecological Applications 13:S185-S198. Levings, C.D. and G.S. Jamieson. 1999. Evaluation of ecological criteria for selecting MPAs in Pacific Region: a proposed semi-quantitative approach. Canadian Stock Assessment Secretariat Research Document 99/210 Levings, C.D., J.M. Helfield, D.J. Stucchi, and T.F. Sutherland. 2002. A perspective on the use of performance based standards to assist in fish habitat management on the seafloor near salmon net pen operations in British Columbia. Canadian Science Advisory Secretariat, research document 2002/75. Levinton, J.S. 1995. Marine biology: function, biodiversity, ecology. Oxford University Press, New York, N.Y. 421 p. Lewis, D.B. and J.J. Magnuson. 1999. Landscape spatial patterns in freshwater snail assemblages across northern highland catchments. Freshwater Biology 41:1-12. Lindenmayer D.B., A.D. Manning, P.L. Smith, H.P. Possingham, J. Fischer, I. Oliver, M.A. McCarthy. 2002. The focal-species approach and landscape restoration: a critique. Conservation Biology 16(2): 338-345 Lindzey, F.G. and E.C. Meslow. 1977. Home range and habitat use by black bears in southwestern Washington. Journal of Wildlife Management 41:413-325. Lloyd, K. and S. Fleck. 1977. Some aspects of the ecology of black and grizzly bears in southeast British Columbia. Unpublished report, BC Fish and Wildlife Branch, Cranbrook, B.C. 55 p. Lombard, A.T., R.M. Cowling, R.L. Pressey, and A.G. Rebelo. 2003. Effectiveness of land classes as surrogates for species in conservation planning for the Cape Floristic Region. Biological Conservation 112:63-85. Lomer, F. and G.W. Douglas. 1998. Additions to the vascular plant flora of the Queen Charlotte Islands, British Columbia. Canadian Field-Naturalist 113:235-240. Longley, P. A., M.F. Goodchild, D.J. Maguire, and D.W. Rhind. 2001. Geographic Information Systems and Science. John Wiley & Sons, Chichester, U.K. 454 p.



Page 172

Lowell, H.S., R.W. Flynn, J.W. Schoen, and L. Shea. 1988. Habitat capability model for mountain goats in southeast Alaska: winter habitat. USDA Forest Service, Alaska Region, Juneau, Alaska. Lyons, J.L. 1989. Correspondence between the distribution of fish assemblages in Wisconsin streams and Omernik’s ecoregions. American Midland Naturalist 122:163-182. Lyon, L.J. 1983. Road density models describing habitat effectiveness for elk. Journal of Forestry 81:592-595. Mace, R.D., J.S. Waller, T.L. Manley, K. Ake, and W.T. Wittinger. 1999. Landscape evaluation of grizzly bear habitat in Western Montana. Conservation Biology 13(2):367-377. Mace, R.D., J.S. Waller, T.L. Manley, L.J. Lyon, and H. Zuuring. 1996. Relationships among grizzly bears, roads, and habitat in the Swan Mountains, Montana. Journal of Applied Ecology 33:1395-1404. MacHutchon, A.G., S. Himmer, and C.A. Bryden. 1993. Khutzeymateen Valley grizzly bear study. Final Report. Wildlife Report R-25. B.C. Ministries of Environment, Lands and Parks and Forests, Victoria, B.C. 105 p. MacKenzie, W.H. and A. Banner. 2000. A classification framework for wetlands and related ecosystems in British Columbia. B.C. Ministry of Forests, Research Branch, Victoria, B.C. MacNally, R., A.F. Bennett, G.W. Brown, L.F. Lumsden, A. Yen, S. Hinkley, P. Lillywhite, and D. Ward. 2002. How well do ecosystem-based planning units represent different components of biodiversity? Ecological Applications 12:900-912. MacPhail, J. and R. Carveth. 1994. Field Key to the Freshwater Fishes of British Columbia. Draft for 1994 field-testing. Fish Museum, Department of Zoology, University of British Columbia. Available online: URL: http://www.for.gov. bc.ca/ric Mahon, T., D. Morgan, and F. Doyle. 2003. Northern Goshawk (Accipiter gentilis) habitat in the North Coast Forest District, foraging area and nest area habitat suitability models (Draft, January 17, 2003). 21 p. Manly, B.F., J.L.L. McDonald, and D.L. Thomas. 1993. Resource selection by animals. Chapman and Hall, New York, NY. Mann, K.H. and J.R.N. Lazier. 1996. Dynamics of marine ecosystems. Biological-physical interactions in the oceans. Blackwell Science, Oxford, U.K. 406 p. Marbled Murrelet Recovery Team. 2003. Marble Murrelet Conservation Assessment Part B: Marble Murrelet Recovery Team advisory document on conservation and management. Unpub. Margules, C.R. and R.L. Pressey. 2000. Systematic conservation planning. Nature 405:243-253. Margules, C.R., A.O. Nicholls, and R.L. Pressey. 1988. Selecting networks of reserves to maximize biological diversity. Biological Conservation 43:63-76.



Page 173

Marine Mammal Center Web page. “Marine Mammal Information: Sea otters.” URL: http://www.marinemammalcenter.org/learning/education/mammalinfo/seaotter.asp (Accessed Dec. 01). Marine Reserves Working Group (MRWG). 2001. Channel Islands Marine Reserves Working Group Science Advisory Panel. How large should marine reserves be? Draft Aug. 30. Mathews, W.J. 1998. Patterns in Freshwater Fish Ecology. Chapman and Hall, New York, N.Y. Maxwell, J.R., C.J. Edwards, M.E. Jensen, S.J. Paustain, H. Parrot, and D.M. Hill. 1995. A hierarchical framework of aquatic ecological units in North America (Nearctic Zone). General Technical Report NC-176. USDA Forest Service, North Central Forest Experimental Station, St. Paul, Minn. McCarter, P.B. and D.E. Hay. 1999. Distribution of spawning eulachon stocks in the Central Coast of British Columbia as indicated by larval surveys. Canadian Stock Assessment Secretariat Research Document 99/177. McCarthy, M.A. 2002. The focal-species approach and landscape modified: a critique. Conservation Biology 16: 338-345. McCrory, W.P., P. Paquet, and B. Cross. 2003. Assessing Conservation for the Gray Wolf and Sitka Deer of the B.C. Central Coast Rainforest (February 2003). Report to the Valhalla Wilderness Society, New Denver, B.C. 80 p. McGurk, B.J. and D.R. Fong. 1995. Equivalent roaded area as a measure of cumulative effect of logging. Environmental Management 19:609-621. Mech, L.D. 1989. Wolf population survival in an area of high road density. Am. Midl. Nat. 121:387-389. Mech, L.D., L.D. Frenzel, Jr., and P.D. Karns. 1971. The effect of snow on the vulnerability of white-tailed deer to wolf predation. Pp. 51-59 in L.D. Mech and L.D. Frenzel, Jr. (editors). Ecological studies of the timber wolf in northeastern Minnesota. USDA For. Serv., Res. Pap. NC-52. Meidinger, D. and J. Pojar (editors). 1991. Ecosystems of British Columbia. B.C. Ministry of Forests, Victoria, B.C. Special Report Series No. 6. 330 p. Metter, D.E. 1968. The influence of floods on population structure of Ascaphus truei Stejneger. Journal of Herpetology 1:105-106. Michelfelder, V. 2003. Effects of partial cutting on understory composition and abundance: emphasizing the Coastal Western Hemlock biogeoclimatic zone. Unpublished report. School of Resources and Environmental Management, Simon Fraser University, Burnaby, B.C. 36 p. Modafferi, R.D. 1978. Black bear movements and home range study. Alaska Department of Fish and Game Report. 23 p.



Page 174

Moore, K. 1991. Coastal watersheds: An inventory of watersheds in the coastal temperate forests of British Columbia. A BC Endangered Spaces Project Working Paper. Earthlife Canada Foundation & Ecotrust/Conservation International, Victoria, B.C. 54 p. Morris, M. and P. Thuringer. 2001. Interactions of coastal species with intertidal habitat types. discussion report. Archipelago Marine Research, Victoria, B.C. Morris, W.F. and D.F. Doak. 2002. Quantitative conservation biology: Theory and practice of population viability analysis. Sinauer Associates, Sunderland, Mass. 480 p. Morrison, P.H. and F. Swanson. 1990. Fire history in two forest ecosystems of the central western Cascade Range, Oregon. General Technical Report. PNW-GTR-254. USDA Forest Service, Portland, Oreg. Mosquin, T., P. G. Whiting, and D. E. McAllister. 1995. Canada's biodiversity: the variety of life, its status, economic benefits, conservation costs, and unmet needs. Canadian Centre for Biodiversity, Ottawa, Ontario. Muhar, S. and M. Jungwirth. 1998. Habitat integrity of running waters - assessment criteria and their biological relevance. Hydrobiologia 386:195-202. Myers, N. 2000. Biodiversity hotspots for conservation priorities. Nature (403):853-858. Naiman, R.J. (editor). 1992. Watershed management: Balancing sustainability and environmental change. Springer-Verlag, New York, N.Y. Naiman, R.J., R.E. Bilby, and P.A. Bisson. 2000. Riparian ecology and management in the Pacific coastal rainforest. BioScience 50(11):996-1011. Naiman, R.J., H. Décamps, and M. Pollock. 1993. The role of riparian corridors in maintaining regional biodiversity. Ecological Applications 3:209-212. National Research Council (NRC). 2000. Marine protected areas: Tools for sustaining ocean ecosystems. National Academy Press, Washington, D.C. National Research Council (NRC). 1996. Upstream, salmon and society in the Pacific Northwest. National Academy Press. Washington, D.C. NatureServe Web page. “NatureServe Explorer: An Online Encyclopedia of Life.” URL: http://www.natureserve.org/explorer/ Newall, P.R. and J.J. Magnuson. 1999. The importance of ecoregion versus drainage area on fish distribution in the St. Croix River and its Wisconsin tributaries. Environmental Biology of Fishes 55:245-254. Newmark, W. D. 1995. Extinction of mammal populations in Western North American national parks. Conservation Biology 9:512-526.



Page 175

Newmark, W. D. 1996. Insularization of Tanzanian parks and the local extinction of large mammals. Conservation Biology 9:512:526. Nicholson D. and J. Booth. 1997. Johnstone Strait and Central Coast Biological and Human Use Mapping Data Dictionary, Clover Point Cartographics, Jacqueline Booth and Associates. Prepared for the Land Use Coordination Office. Noss, R. 1983. A regional landscape approach to maintaining diversity. Bioscience 33:700-706. Noss, R.F. 1987. From plant communities to landscapes in conservation inventories: A look at The Nature Conservancy. Biological Conservation 41:11-37. Noss, R.F. 1990. Indicators for monitoring biodiversity: A hierarchical approach. Conservation Biology 4(4):335-364. Noss, R.F. 1991. From endangered species to biodiversity. Island Press, Washington, D.C. Noss, R.F. 1993. A conservation plan for the Oregon coast range: Some preliminary suggestions. Natural Areas Journal 13(4):276-290. Noss, R.F. 1996. Protected areas: How much is enough? Pp. 91-120 in R.G. Wright (editor). National Parks and Protected Areas. Blackwell, Cambridge, Mass. Noss, R. 2001. Beyond Kyoto: Forest management in a time of rapid climate change. Conservation Biology 15:578-590. Noss, R.F. 2003. A checklist for wildlands network designs. Conservation Biology 17(5):1270-1275. Noss, R.F., C. Carroll, K. Vance-Borland, and G. Wuerthner. 2002. A multi-criteria assessment of the irreplaceability and vulnerability of sites in the Greater Yellowstone Ecosystem. Conservation Biology 16(4):895-908. Noss, R.F. and A.Y. Cooperrider. 1994. Saving nature’s legacy: Protecting and restoring biodiversity. Island Press, Washington, D.C. 416 p. Noss, R.F., M.A. O’Connell, and D. D. Murphy. 1997. The science of conservation planning: Habitat conservation under the Endangered Species Act. Island Press, Washington, D.C. and Covelo, Calif. Noss, R.F., J.R. Strittholt, K. Vance-Borland, C. Carroll, and P. Frost. 1999. A conservation plan for the Klamath-Siskiyou ecoregion. Natural Areas Journal 19:392-411. Odum, E.P. 1970. Optimum population and environment: a Georgia microcosm. Current History 58: 355-359. Odum, E.P. and H.T. Odum. 1972. Natural areas as necessary components of man’s total environment. Proc. N.A. Wildl. Nat. Res. Conf. 37:178-189. Oliver, C.D. and B.C. Larson. 1990. Forest stand dynamics. McGraw-Hill, New York, NY.



Page 176

Pacific Estuary Conservation Program. 2003. Pacific Estuary Conservation Program (PECP) British Columbia estuary mapping project. Digital datafiles. Pacific States Marine Fisheries Commission Web page. Eulachon Fact Sheet. URL: http://www.psmfc.org/habitat/edu_smelt_fact.html (accessed Dec. 01). Panjabi, A. (editor). 2001. The partners in flight handbook on species assessment and prioritization version 1.1, December 2001. Prepared for Partners In Flight. Available online: URL: http://www.rmbo.org/pif/pifdb.html Panzer, R. and M.W. Schwartz. 1998. Effectiveness of a vegetation-based approach to conservation. Conservation Biology 12:693-702. Paquet, P.C. and C. Callaghan. 1996. Effects of linear developments on winter movements of gray wolves in the Bow River Valley of Banff National Park, Alberta. Pp. 1-21 in G.L. Evink, P. Garret, D. Zeigler, and J. Berry (editors). Proceedings of the International Conference on Wildlife Ecology and Transportation, Tallahassee, Fla. Parks, S.A. and A.H. Harcourt. 2002. Reserve size, local human density and mammalian extinction in U.S. protected areas. Conservation Biology 16(3):800-808. Parks Canada. 1999. Representing the Queen Charlotte Sound Marine Region in Parks Canada’s National Marine Conservation System. Peek, J. M. 2000. Chapter 23: Mountain Goat in S. Demarais, and P. R. Krausman (editors). Ecology and management of large mammals in North America. Prentice-Hall, Inc. Peek, J., J. Beecham, D. Garshelis, F. Messier, S. Miller. and D. Strickland. 2003. Management of grizzly bears in British Columbia: A review by an independent scientific panel. B.C. Ministry of Water, Land and Air Protection, Victoria, B.C. 90 p. Pelton, M.R. 1982. Black bear. Pp. 504-514 in J.A. Chapman and G.A. Feldhamer (editors). Wild Mammals of North America. John Hopkins University Press, Baltimore, Md. Perry, I.R., M. Stocker, and J. Fargo. 1994. Environmental effects on the distributions of ground fish in Hecate Strait, British Columbia. Canadian Journal of Fisheries and Aquatic Sciences 51(6):1401-1409. Peterman, R.M. 1990. Statistical power analysis can improve fisheries research and management. Canadian Journal of Fisheries and Aquatic Sciences 47:2-15. Peterson, G., C.R. Allen, and C. S. Holling. 1998. Ecological resilience, biodiversity and scale. Ecosystems 1:6-18. Pickett, S.T.A. and J. Thompson. 1988. Patch dynamics and the design of nature reserves. Biological Conservation 13:27-37.



Page 177

Pimentel, D., L. Westra, and R. Noss (editors). 2000. Ecological integrity. Integrating environment, conservation, and health. Island Press, Washington, D.C. 428 p. Poff, N.L and J.V. Ward. 1989. Implications of streamflow variability and predictability for lotic community structure and a regional analysis of streamflow pattern. Canadian Journal of Fisheries and Aquatic Science 46:1805-1818. Poff, N.L. and J.D. Allan. 1995. Functional organization of stream fish assemblages in relation to hydrologic variability. Ecology 76:606-627. Pojar, J. and A. MacKinnon (editors). 1994. Plants of the Pacific Northwest coast. Lone Pine, Vancouver, B.C. Pojar, J., K. Klinka, and D.V. Meidinger. 1987. Biogeoclimatic ecosystem classification in British Columbia. Forest Ecology and Management 22:119-154. Pollard, B. 2002a. Draft environmental risk assessment base case: mountain goats. Acer Resource Consulting, Ltd. Pollard, B. 2002b. Mountain goat winter range mapping for the North Coast Forest District. North Coast Land and Resource Management Planning Team, Victoria, B.C. Poole, K.G. and G. Mowat. 1997. Mountain goat winter habitat use in the West Kootenay region of British Columbia. Timberland Consulting Ltd, Nelson, B.C. Possingham, H.P., I. Ball, and S. Andelman. 2000. Mathematical methods for identifying representative reserve networks. Pp. 291-306 in S. Ferson and M.A. Burgman (editors). Quantitative methods in conservation biology. Springer-Verlag, New York, N.Y. Power, M.E., D. Tilman, J.A. Estes, B.A. Menge, W.J. Bond, L.S. Mills, G. Daily, J.C. Castilla, J. Lubchenco, and R.T. Paine. 1996. Challenges in the quest for Keystones. BioScience 46(8):609-620. Pressey, R.L., H.P. Possingham, and C.R. Margules. 1996. Optimally in reserve selection algorithms: when does it matter and how much? Biological Conservation 76:259-267. Pressey, R.L. 1994. Land classifications are necessary for conservation planning, but what do they tell us about fauna? Pp. 31-41 in D. Lunney, S. Hand, P. Reed, and D. Butcher (editors). Future of the fauna of western New South Wales. Surrey Beatty and Sons, Chipping Norton. Pressey, R.L. and V.S. Logan. 1994. Level of geographic subdivision and its effects on assessments of reserve coverage: a review of regional studies. Conservation Biology 8:1037-1046. Pressey, R.L., I.R. Johnson, and P.D. Wilson. 1994. Shades of irreplaceability: Towards a measure of the contribution of sites to a reservation goal. Biodiversity and Conservation 3:242-262. Province of B.C. 1995. Conservation of Grizzly Bears in British Columbia. Background report. B.C. Ministry of Environment, Lands and Parks, Victoria, BC. 70 pp.



Page 178

Quarles, H.D., R.B. Hanawalt, and W.E. Odum. 1974. Lead in small mammals, plants and soil at varying distances from a highway. Journal of Applied Ecology 11:937-949. Redford, K.H. and B.D. Richter. 1999. Conservation of biodiversity in a world of use. Conservation Biology 13: 1246–1256.

Reimchen, T.E. 1998a. Nocturnal foraging behaviour of Black bear, Ursus americanus, on Moresby Island, British Columbia. Canadian Field-Naturalist 112:446-450 Reimchen, T.E. 1998b. Diurnal and nocturnal behaviour of Black bears, Ursus americanus, on bear trails. Canadian Field-Naturalist 112:698-699 Remington, D. 1993. Coastal wetlands habitat assessment and classification for northwestern British Columbia. North Coast Wetlands Program. For: Pacific Estuary Conservation Program. Resources Inventory Committee (RIC). 1999. British Columbia wildlife habitat ratings standards. Version 2.0. B.C. Ministry of Environment, Lands and Parks, Victoria, B.C. 97 p. Resources Inventory Committee (RIC). 2000. Standards for broad terrestrial ecosystem classification and mapping for British Columbia. Version 2.0. Ecosystems Working Group, Terrestrial Ecosystems Task Force, Victoria, B.C. 212 p. Reyers B, A.S. van Jaarsveld, and M. Krüger. 2000. Complementarity as a biodiversity indicator strategy. Proceeding of the Royal Society of London Series B – Biological Sciences 267(1442) Ricketts, T.H., E. Dinerstein, D.M. Olson, C.J. Loucks, W. Eichbaum, D. DellaSalla, K. Kavanaugh, P. Hedao, P. Hurley, K. Carney, R. Abell, and S. Walters. 1999. Terrestrial ecoregions of North America: a conservation assessment. Island Press, Washington, D.C. Ritland, K., C. Newton, and H.D. Marshall. 2001. Inheritance and population structure of the white-phased Kermode Black bear. Current Biology 11:1468-1472 Roberts, C.M. and J.P. Hawkins. 2000. Fully protected marine reserves: a guide. WWF Endangered Seas Campaign, Washington, D.C., and Environment Department, University of York, York, U.K. Roberts, C.M., S. Andelman, G. Branch, R.H. Bustamante, J.C. Castilla, J. Dugan, B.S. Halpern, K.D. Lafferty, H. Leslie, J. Lubchenco, D. McArdle, H. Possingham, M. Ruckelshaus, and R. Warner. 2003. Ecological criteria for evaluating candidate sites for marine reserves. Ecological Applications 13:1 supplement 199-214. Robinson, C. 2002. Species at risk newsletter no. 18: Parks Canada December 2002, Critical marine foraging habitats for Marbled Murrelets in Pacific Rim National Park Reserve of Canada. Parks Canada. Vancouver, BC. 2 p. Rodrigues, A.S.L. and K.J. Gaston. 2001. How large do reserve networks need to be? Ecology Letters 4(6):602-609.



Page 179

Rodrigues, A.S.L., R. Tratt, B.D. Wheeler, and K.J. Gaston. 1999. The performance of existing networks of conservation areas in representing biodiversity. Proceedings of the Royal Society of London Series B - Biological Sciences 266 (1427):1453-1460. Roff, J.C. and S. Evans. 2001. Viewpoint: National frameworks for marine conservation -non-hierarchical approaches and distinctive areas. Unpublished. Rogers, L.L. 1977. Social relationship, movements, and population dynamics of black bears in northeastern Minnesota. Ph.D. thesis, Univ.Minn., Minneapolis, Minn. 194 p. Rogers, L.L. and AW. Allen. 1987. Habitat suitability index model: black bear. Upper Great Lakes Reg. USFW Serv.Biol.Rep. 82. 54 p. Rost, G.R. and J.A. Bailey. 1979. Distribution of mule deer and elk in relation to roads. Journal of Wildlife Management 43(3):634-641. Roth, N.E., J. D. Allan, and D.L. Erickson. 1996. Landscape influences on stream biotic integrity assessed at multiple scales. Landscape Ecology 11:141-156. Ruff, R.L. 1982. Dynamics of black bear populations: (low to no human exploitation). Pp. 87-103 in F.G. Lindzey (editor). Procedures of the Second West Black Bear Workshop, Logan, Utah. Rumsey, C., M. Wood, B. Butterfield, P. Comer, D. Hillary, M. Bryer, C. Carroll, G. Kittel, K.J. Torgerson, C. Jean, R. Mullen, P. Iachetti, and J. Lewis. 2003. Canadian Rocky Mountains Ecoregional Assessment, Volume One: Report. Prepared for The Nature Conservancy and the Nature Conservancy of Canada. Victoria, BC. 88 p. Available online: URL: http://www.natureconservancy.ca/pdf/crm_vol1_report.pdf Russell, D. 1974. Grizzly bear-mountain goat investigations in Knight Inlet, British Columbia. B.C. Fish and Wildlife Branch, Nanaimo, B.C. 84 p. Ryti, R. T. 1992. Effects of the focal taxon on the selection of nature reserves. Ecological Applications 2:404-410. Savard, J-P.L. 1988. A summary of current knowledge on the distribution and abundance of moulting seaducks in the coastal waters of British Columbia. Canadian Wildlife Service, Pacific and Yukon region. Technical Report Series No. 45. Schmidt, D.C., S.R. Carlson, G.B. Kyle, and B.P. Finney. 1998. Influence of carcass-derived nutrients on sockeye salmon productivity of Karluk Lake, Alaska: Importance in the assessment of an escapement goal. N. Amer. J. Fish. Man. 18:743-763. Schoen, J.W. and M.D. Kirchhoff. 1990. Seasonal habitat use by Sitka black-tailed deer on Admiralty Island. Journal of Wildlife Management 54:371-378. Schoen, J.W. and M.D. Kirchhoff. 1982. Habitat use by mountain goats in southeast Alaska. Page 67. Alaska Department of Fish and Game; Federal Aid in Wildlife Modified Project W-17-10, 11 and W-21-1, 2; Final Report; job 12.4R, Juneau, Alaska.



Page 180

Schoen, J.W., M.D. Kirchhoff, and J.H. Hughes. 1988. Wildlife and old-growth forests in southeastern Alaska. Natural Areas Journal 8:138-145. Schoen, J.W., M.D. Kirchhoff, and O.C. Walmo. 1980. Winter habitat use by mountain goats. Page 67. Alaska Department of Fish and Game; Federal Aid in Wildlife Modified Project W-17-10, 11 and W-21-1, 2; Final Report; job 12.4R, Juneau, Alaska. Schoonmaker, P.K., B. von Hagen, and E.C. Wolf. 1997. The rain forests of home: Profile of a North American bioregion. Island Press, Washington, D.C. Scott, J.M. 1999. A representative biological reserve system for the United States? Society for Conservation Biology Newsletter 6(2):1,9. Scott, J.M., F. Davis, B. Csuti, R. Noss, B. Butterfield, C. Groves, J. Anderson, S. Caicco, F. D’Erchia, T.C. Edwards, J. Ulliman, and R.G. Wright. 1993. Gap analysis: a geographical approach to protection of biological diversity. Wildlife Monographs 123:1-41. Scudder, G.G.E. 1996. Terrestrial and freshwater invertebrates of British Columbia: Priorities for inventory and descriptive research. B.C. Ministry of Forests Research Program, Victoria, B.C. 206 p. Searing, G.F. and H.R. Frith. 1995. British Columbia shore-zone mapping system. Contract report by LGL Ltd., Sidney, B.C for the Land Use Coordination Office, B.C. Ministry of Environment, Victoria, B.C. 46 p. Severinghaus, C.W. 1947. Relationship of weather to winter mortality and population levels among deer in the Adirondack region of New York. Transactions of the North American Wildlife Conference 12:212-223. Shaffer, M.L. 1981. Minimum population sizes for conservation. BioScience 31: 131-134. Sharpe, S.T., A.J. Gaston, and J.L. Martin. 2002. Forest ecology, forest renewal and introduced species in Haida Gwaii, Final Report. Prepared for Science Council of British Columbia Forest Renewal BC Program, Burnaby, B.C. 39 p. Shrader-Frechette, K.S. and E.D. McCoy. 1993. Method in ecology: Strategies for conservation. Cambridge University Press, New York, NY. Shugard, H. and D. West. 1981. Long-term dynamics of forest ecosystems. American Scientist 69:647-652. Sims, R. 1999. Terrestrial Ecosystem Mapping (TEM) with wildlife habitat interpretations of the Besa-Prophet area: Part 2: Wildlife Report. Page 142. R.A. Sims and Associates, prepared for B.C. Environment, Lands and Parks Oil and Gas Diversion, Vancouver, B.C. Smith, B.L. 1977. Influence of snow conditions on winter distribution, habitat use, and groups size of mountain goats. Pp. 174-189 in W. Sammuel and W.G. Macgregor (editors). Proceedings, First international mountain goat conference. Province of British Columbia Ministry of Recreation and Conservation, Fish and Wildlife Branch, Kalispell, Mont.



Page 181

Smith, C.A. 1986. Habitat use by mountain goats in southeastern Alaska. Alaska Dep. Fish and Game. Fed Aid in Wildl. Rest. Final Report. Smith, C.A. 1994. Evaluation of a multivariate model of mountain goat winter habitat selection. Pp. 159-165 in M. Pybus (editor). Proceedings of the ninth biennial symposium of the Northern Wild Sheep and Goat Council, Cranbrook, BC. Smith, W. and P. Lee. 2000. Canada’s forests at a crossroads: an assessment in the year 2000. World Resources Institute, Washington D.C. Soulé, M.E. and M.A. Sanjayan. 1998. Conservation targets: Do they help? Science 279:2060-2061. Soule, M. E., and D. Simberloff. 1986. What do genetics and ecology tell us about the design of nature reserves? Biological Conservation 35:19-40. Spies, T. and J.F. Franklin. 1995. The diversity and maintenance of old growth forests. Pp. 296-314 in Biodiversity in managed landscapes, R.C. Szaro and D.W. Johnson (editors). Oxford Univ. Press, New York, N.Y. Squires, J.R. and R.T. Reynolds. 1997. Northern Goshawk (Accipiter gentilis). In A. Poole and F. Gill (editors). The birds of North America, No. 298. Academy of Natural Sciences, Philadelphia, Penn., and American Ornithologists’ Union, Washington, D.C. 32pp. Stanford, J.A. and J.V. Ward. 1992. Management of aquatic resources in large catchments: Recognizing interactions between ecosystem connectivity and environmental disturbance. Pp. 91-124 in R.J. Naiman (editor). Watershed management. Springer-Verlag, New York, N.Y. Stevens, V. and S. Lofts. 1988. Wildlife habitat handbooks for the Southern Interior Ecoprovince Vol. 1: Species Notes for Mammals. BC Environ., BC Min. For., Victoria, B.C. Steventon, J.D. 2002. Environmental Risk Assessment Base Case: Marbled Murrelet (Draft 16/09/02). Prepared for North Coast LRMP. 35 pp. Stiassny, M.L. 1996. An overview of freshwater biodiversity. Fisheries 21: 7-13. Stoddard, M.A. 2002. The influence of forest management on headwater stream amphibians at multiple spatial scales. MSc. thesis, Oregon State University, Corvallis, Oreg. Suring, L.H., E.J. Degayner, R.W. Flynn, M.D. Kirchhoff, J.W. Schoen, and L.C. Shea. 1992. Habitat Capability Model for Sitka Black-tailed Deer in Southeast Alaska: Winter Habitat (Version 6.5, April 1992). USDA Forest Service, Region 10. 95 p. Suring, L.H., R.W. Flynn, J.W. Schoen, and L.C. Shea. 1988. Habitat capability model for mountain goats in Southeast Alaska: Winter habitat. USDA Forest Service, Alaska Region, Juneau, Alaska. Sustainable Ecosystems Institute (SEI) Web page. “Endangered Species: Marbled Murrelet.” URL: http://www.sei.org/murrelet.html (accessed Sept. 2003).



Page 182

Sutherland, G.D., M.P. Hayes, T. Quinn, L.A. Dupuis, T.R. Wahbe, D.E. Runde, and J.S. Richardson. 2001. Predictive habitat models for the occurrence and abundance of the Olympic Tailed Frog, and the Rocky Mountain Tailed Frog, A pilot meta-analysis. Report for the Landscape and Wildlife Advisory Group and the Amphibian Research Consortium submitted to the Washington State Department of Natural Resources, Olympia, Wash. Swetnam, T.W., C.D. Allen, and J.L. Betancourt. 1999. Applied historical ecology: Using the past to manage the future. Ecological Applications 9:1189-1206. Taylor, B.L. and T. Gerrodette. 1993. The uses of statistical power in conservation biology: the vaquita and the Northern Spotted Owl. Conservation Biology 7:489-500. The Nature Conservancy (TNC). 1996. Conservation by design: A framework for mission success. Arlington, Va. 16 p. The Nature Conservancy (TNC). 1997. Designing a geography of hope: Guidelines for ecoregion based conservation in The Nature Conservancy. Arlington, Va. 84 p. The Nature Conservancy (TNC). 1999. Geography of hope: Practical guidelines for setting conservation priorities in ecoregions. Review draft. Arlington, Va. 113 p. plus appendices. The Nature Conservancy (TNC). 2000. Designing a geography of hope: A practitioner’s handbook to ecoregional conservation planning. Arlington, Va. Thomson, R.E. 1981 (Reprinted 1983, 1984, 1991) Oceanography of the British Columbia Coast. Can. Special Pub. Fish. Aquat. Sci. 56. Ottawa, Ont. Thorne-Miller, B. 1999. The living ocean. Understanding and protecting marine biodiversity. 2nd ed. Island Press, Washington, DC. Tonn, W.M. 1990. Climate change and fish communities: a conceptual framework. Transactions of the American Fisheries Society 119:337-352. Unsworth, J.W., J.J. Beecham, and L.R. Irby. 1989. Female black bear habitat use in west-central Idaho. Journal of Wildlife Management 53:668-673. U.S. Fish and Wildlife Service. 1981. Standards for the development of habitat suitability index models. Ecological Services Manual 103. Dept of the Interior. Washington, D.C. U.S. Fish and Wildlife Service. 1993. Grizzly bear recovery plan. U.S. Fish and Wildlife Service Final Report. 181 p. Utzig, G.F. and T.M. Gaines. 1998. Application of a Habitat Suitability Index for the Northern Goshawk in the SIMFOR Habitat Model. Report prepared for B.C. Environment. 23 p. Van Dyke, F.B., R.H. Brocke. H.G. Shaw, B.B. Ackerman, T.P. Hemker and F.G. Lindzey. 1986. Reactions of mountain lions to logging and human activity. Journal of Wildlife Management 50: 95-102.



Page 183

Vannote, R.L., G.W. Minshall, K.W. Cummins, and J.R. Sedell. 1980. The river continuum concept. Canadian Journal of Fisheries and Aquatic Science 37:130-137. Welsh, H.H., Jr. and A.J. Lind. 1991. The structure of the herpetofaunal assemblage in the Douglas-fir/hardwood forests of northwestern California and southwestern Oregon. Pp. 395-411 in L.F. Ruggiero, K.B. Aubry, A.B. Carey, and M.H. Huff (technical coordinators). Wildlife and vegetation of unmanaged Douglas-fir forests. USDA Forest Service, Gen. Tech. Rep. PNW-GTR-285. Welsh, H.H., Jr. and A.J. Lind. 2002. Stream amphibians of the Klamath-Siskiyou region: relationships with forest and stream environments. Journal of Wildlife Management 66:581-602. Wessels, KJ, S. Freitag, and A.S. van Jaarsveld. 1999. The use of land facets as biodiversity surrogates during reserve selection at a local scale. Biological Conservation 89:21–38. West, J.E. 1997. Protection and modified of marine life in the inland marine waters of Washington State. Report prepared for the Puget Sound/Georgia Basin International Task Force, Washington Work Group on Protecting Marine Life. 144 p. Whittaker, R.H. 1975. Communities and ecosystems. 2nd ed. Macmillan, New York, NY. Wielgus, R. B. 2002. Minimum viable population and reserve sizes for naturally regulated grizzly bears in British Columbia. Biological Conservation IN PRESS. Wigal, R.A. and V.L. Coggins. 1982. Mountain goat Oreamnos americanus. Pp. 1008-1020 in J.A. Chapman and G. A. Feldhamer (editors). Wild mammals of North America. Johns Hopkins University Press, Baltimore, Md. Williams, P.H. and C.J. Humphries. 1996. Comparing character diversity among biotas. Pp. 54-76 in K.J. Gaston (editor). Biodiversity: A biology of numbers and difference. Blackwell Science. Oxford, U.K. Wilson, M.F. and K.C. Halupka. 1995. Anadromous fish as keystone species in vertebrate communities. Conservation Biology 9(3): 489-497. Woodroffe, R., and J. R. Ginsberg. 1998. Edge effects and the exinction of populations inside protected areas. Science 280:2126-2128. Yost, A.C. and R.G. Wright. 2001. Moose, caribou and grizzly bear distribution in relation to road traffic in Denali National Park, Alaska. Arctic 54(1):41-48. Young, B.F. and R.L. Ruff. 1982. Population dynamics and movements of black bears in east central Alberta. Journal of Wildlife Management 46:845-860. Zacharias, M.A. and J.C. Roff. 2000. A hierarchical ecological approach to conserving marine biodiversity. Conservation Biology 14(5):1327-1334.



Page 184

Zacharias, M.A. and J.C Roff. 2001. Use of focal species in marine conservation and management: a review and critique. Aquatic Conservation: Marine and Freshwater Ecosystems 11:59-76. Zeiler, M. 1999. Modeling our world: The ESRI guide to geodatabase design. Environmental Systems Research Institute, Inc. 199 p.



Page 185

GLOSSARY

Biodiversity: The variety of living organisms, the ecological complexes in which they occur, and the ways in which they interact with each other and the physical environment It characterizes biodiversity as having three primary components: composition, structure, and function. From a conservation perspective, it is necessary to consider each of these components. Community: An interacting assemblage of species that co-occur with some degree of predictability and consistency. Critical wildlife habitat: Part or all of a specific place occupied by a wildlife species population of such species and recognized as being essential for the maintenance of the population: wetlands, breeding sites, mineral licks, rutting arenas, etc.), birthing sites (calving, spawning, etc.), riparian zones, colonies, rookeries, hibernacula, winter range and over wintering area (caribou, ungulates, trumpeter swan, etc.), caves, talus slopes, avalanche chutes, denning sites, nesting sites, and cliffs. Ecosystem: A functional unit consisting of all the living organisms (plants, animals, and microbes) in a given area, and all the non-living physical and chemical factors of their environment, linked together through nutrient cycling and energy flow. An ecosystem can be of any size—a log, pond, field, forest or the earth’s biosphere—but it always functions as a whole unit. Ecosystems are commonly described according to the major type of vegetation, for example, forest ecosystem, or range ecosystem. Ecological integrity: The soundness or wholeness of the processes and organisms composing the ecosystem. To maintain ecological integrity one must maintain functioning, self-sustaining ecosystems with characteristics similar to the original ones. Habitat: The place where an organism lives and/or the conditions of that environment including the soil, vegetation, water, and food. Habitat capability: The ability of the habitat, under optimal conditions to provide life requisites of a species, irrespective of its current habitat conditions. It is the potential of a forested ecosystem under ideal conditions for the wildlife species in question to support that species - the right plant species, forest stand age, and stocking rate for each wildlife species. The current stage age is not relevant to a capability assessment, because what is being evaluated is the site series, with the assumption that at some stage it can produce the forage that is required to support the species in question. All forests go through changes, and once a forest has been logged or burned there is a series of well-defined stages: herb & low shrub, tall shrub, pole sapling, young forest, mature forest, and old forest. If one of those stages has the ability to provide the necessary forage, then the site series is rated for that value for the species Habitat effectiveness: A measure of a species ability to use current or potential habitat conditions. It is an estimate of the effects of human activities, such as roads, fences, recreational uses, industrial developments, settlements, on the usability of the habitat. Habitat suitability: The ability of the habitat in its current condition to provide the life requisites of a species. It is the potential of a forested ecosystem in its current state to support a given



Page 186

wildlife species - the current plant species, current stand age and current stocking rate. It is an estimate of how well current habitat conditions provide the specified life requisites of the species being considered. The suitability of the land is frequently less than the capability because of unfavourable seral conditions or conflicting land use. Historical Area: Refers to the total extent of an ecosystem target regardless of logging history. Keystone species: A species that plays an important ecological role in determining the overall structure and dynamic relationships within a biotic community. A keystone species’ presence is essential to the integrity and stability of a particular ecosystem.

Natural: Indicates a relatively low level of influence by modern human activities. Old growth: Forest that contains live and dead trees of various sizes, species, composition, and age-class structures. Old-growth forests, as part of a slowly changing but dynamic ecosystem, include climax forests but not sub-climax or mid-seral forests. The age and structure of old growth varies significantly by forest type and from one biogeoclimatic zone to another. Used as a convenient shorthand for the important but complex structural and compositional aspects of late-seral forest stands.

Old growth attributes: Structural attributes and other characteristics of old growth forests, including: large trees for the species and site; wide variation in tree sizes and spacing; accumulations of large dead standing and fallen trees; multiple canopy layers; canopy gaps and understory patchiness; elements of decay such as broken or deformed tops or trunks and root decay; and the presence of species characteristic of old growth. Riparian: The land adjacent to the normal high water line in a stream, river or lake and extending to the portion of land that is influenced by the presence of the adjacent ponded or channelled water. Riparian areas typically exemplify a rich and diverse vegetative mosaic reflecting the influence of available surface water.

Seral stages: The stages of ecological succession of a plant community (e.g., from young stage to old stage. The characteristic sequence of biotic communities that successively occupy and replace each other by which some components of the physical environment become altered over time. Structural attributes: Components of a forest stand (including living and dead standing trees, canopy architecture, and fallen dead trees) which together determine stand structure. Watershed: An area of land that collects and discharges water into a single main stream through a series of smaller tributaries.



Page 187

MAPS (attached separately at http://citbc.org/anaecos.html)

Map 1. Central Coast, North Coast, and Haida Gwaii LRMP Boundaries

Map 2. Ecosections and Terrestrial Stratification Units

Map 3. Ecological Drainage Units

Map 4. Conservation Data Centre Element Occurrence Records

Map 5. Forested and Developed Landscape

Map 6a. Biogeoclimatic Ecosystem Classification

Map 6b. Seral Stage and Site Index

Map 7. Focal Ecosystems

Map 8. Habitat Suitability Model for Marbled Murrelet

Map 9. Nesting and Foraging Habitat Suitability Models for Northern Goshawk

Map 10. Mountain Goat Winter Range Habitat Model

Map 11. Grizzly Bear and Black Bear Habitat Capability

Map 12. Freshwater Ecosystems

Map 13. Habitat Suitability Model for Tailed Frog

Map 14. Steelhead Population Status: Winter Run

Map 15. Steelhead Population Status: Summer Run

Map 16. Salmon Population Status: Coho

Map 17. Salmon Population Status: Chinook

Map 18. Salmon Population Status: Chum

Map 19. Salmon Population Status: Even Year Pink

Map 20. Salmon Population Status: Odd Year Pink

Map 21. Salmon Population Status: Sockeye

Map 22. Nearshore Ecosection Stratifications

Map 23. Nearshore Cost Parameters

Map 23a. Shoreline Cost Index

Map 24. Nearshore Data Collection Regions

Map 25. Seabird Colonies

Map 26. Moulting Seaducks



Page 188

Map 27. Anadromous Streams

Map28. Bathymetry

Map 29. Benthic Topographical Complexity

Map 30. Ecological Regions and Data Regions

Map 31. Human Impacts Assessment

Map 32. Conservation Utility of CIT Analysis Units: Existing protected areas locked in.

Map 33. Conservation Utility of CIT Analysis Units: Existing and Candidate protected areas locked in.

Map 34. Conservation Utility of CIT Analysis Units: Existing, Candidate, and Option protected areas locked in.

Map 35. Conservation Utility of CIT Analysis Units: No areas locked in.

Map 36a. Tier One Analysis Units for CIT ESA: Existing protected areas locked in.

Map 36b. Conservation Priority for CIT ESA Analysis Units: Existing protected areas locked in.

Map 37a. Tier One Analysis Units for CIT ESA: Existing and Candidate protected areas locked in.

Map 37b. Conservation Priority for CIT ESA Analysis Units: Existing and Candidate protected areas locked in

Map 38a. Tier One Analysis Units for CIT ESA: Existing, Candidate and Option protected areas locked in.

Map 38b. Conservation Priority for CIT ESA Analysis Units: Existing, Candidate and Option protected areas locked in.

Map 39a. Tier One Analysis Units for CIT ESA: No areas locked in.

Map 39b. Conservation Priority for CIT ESA Analysis Units: No areas locked in.

Map 40. Relative Biodiversity Index for Coastal LRMP’s.

Map 41. Neashore Summed Solution Gradient

Map 42. Nearshore Option A

Map 43. Nearshore Option B

Map 44. Nearshore Option C

Map 45. Deepwater Marine Conservation Utility


Ecosystem Spatial Analysis APPENDICES

April 2004

APPENDICES (attached separately at http://citbc.org/anaecos.html)

1 CONSERVATION TARGETS ............................................................................................................................ 1 1.1.1 Special Elements ............................................................................................................................. 1 1.1.2 Focal species.................................................................................................................................... 8

1.1.2.1 Northern Goshawk (Accipiter gentilis) Nest and Forage Area Suitability Ratings.......... 8 1.1.2.2 Grizzly Bear (Ursus Arctos) Broad Ecosystem Unit Ratings Table .................................. 10 1.1.2.3 Summary of focal species models reviewer comments...................................................... 27

2 NEARSHORE MARINE ................................................................................................................................... 50

2.1 Technical Workshop ............................................................................................................................... 50 2.2 Definitions of the B.C. Representative Shoreline Types..................................................................... 51

3 OFFSHORE MARINE ....................................................................................................................................... 53

3.1 CIT Marine ESA Feature Layers............................................................................................................ 53 3.2 Stream Richness x Magnitude ............................................................................................................... 58

4 HUMAN IMPACTS .......................................................................................................................................... 62

4.1 Coastal Temperate Rainforest Watershed Units ................................................................................. 62 5 RESULTS ............................................................................................................................................................ 64

5.1 Results from the analysis of the CC LRMP final agreement of December 12, 2003 ....................... 64 5.2 Gains achieved for focal ecosystems, old-growth ecosystems, and freshwater systems............... 65

coast information team - british columbia · coast information team ... rick tingey, ken...

Documents